.

 

 

Prepared by :

Aimagin’s Application Team

Project Advisor :

Cdr.Dr. Krisada Sangpetchsong

Aimagin CO.,LTD

Electronic Muscle Stimulation

Electronic Muscle Stimulation

นายจาวกรณ ฉนทนะสขศลป

วศวกรรมไฟฟา จฬาลงกรณมหาวทยาลย

นายธาดา จระจรส

วศวกรรมชวการแพทย มหาวทยาลยมหดล

โครงงานฉบบนสงมอบใหแก

กตตกรรมประกาศ

โครงงานนไมอาจสาเรจได หากขาดความชวยเหลอจาก นาวาโท ดร.กฤษฏา แสงเพชรสอง ซงเปนท

ปรกษาโครงงานในครงน พวกผมรสกซาบซงในความอนเคราะหของดร.กฤษฏา และขอขอบคณอยางยงท

ใหคาแนะนาดๆกบพวกผม

ขอกราบขอบพระคณคณขนษฐา ประสารสขหรอพแดง ทคอยชวยเหลอประสานงานอานวยความ

สะดวกในเรองตางๆ การทากจกรรมตางๆ รวมทงวธการใชชวตในสงคมกบผอนดวยครบ

ขอขอบคณพโกและพโอตจากภาควชาชวะการแพทย มหาวทยาลยมหดล ทคอยใหคาแนะนาในเรอง

วธการทาฮารตแวรตางๆ รวมถงอาจารยเซง เลศมโนรตนทใหพวกผมไดใชสถานทในการทางานทางดาน

ฮารตแวรดวยครบ

ขอขอบคณเพอนๆในคายไฮเปอรแคมป 54 ทกๆคน ทคอยใหคาสนบสนน คาแนะนา และกาลงใจ

ตางๆตลอดระยะเวลา 2 เดอนทฝกงาน

สดทายนขาพเจาขอกราบขอบพระคณ บดา มารดา และครอบครวของขาพเจาทเปนกาลงใจ และให

การสนบสนนชวยเหลอในทกๆเรอง

คณคาและประโยชนอนพงมาจากรายงานเลมน ขาพเจาขอมอบใหแกผมพระคณทกทาน

นายจาวกรณ ฉนทนะสขศลป

นายธาดา จระจรส

 

          

-A-

บทคดยอ

โครงงานนเปนโครงงานทไดรบมอบหมายมาใหสรางระบบประยกตโดยใชโปรแกรม Simulink ซง

ระบบประยกตทพวกผมเลอกทากคอเครองกระตนกลามเนอ (Electronic Muscle Stimulation )

โครงงานนประกอบดวยทงสวนของซอฟตแวรและฮารตแวร โดยการทางานของเครองจะแบง

ออกเปน 2 สวนหลกๆ กคอสวนสาหรบผปวยกลามเนอออนแรง ในสวนนผใชสามารถปรบคาความแรงได

เองตามตองการ และสวนสาหรบผปวยโรคเทาตก ในสวนนพวกผมไดทาการคงคาความแรงไวเปนคาๆหนง

ทเหมาะสมกบผใชทวๆไป

เนองจากพวกผมมเวลาในการทาโครงงานทจากด พวกผมจงไมไดพฒนาในสวนของหนวยความจา

ททาใหเครองสามารถเกบขอมลตางๆไวไดตามทอาจารยทานตางๆไดแนะนามา

-B-

สารบญ

กตตกรรมประกาศ_____________________________________________________________________ A

บทคดยอ_____________________________________________________________________________ B

สารบญ______________________________________________________________________________ C

สารบญภาพ___________________________________________________________________________ F

บทท 1 บทนา__________________________________________________________________________ 1

1.1 ทมาและภาพรวมของโครงงาน __________________________________________________ 1

1.2 วตถประสงคของโครงงาน______________________________________________________ 2

1.3 ประโยชนทคาดวาจะไดรบ _____________________________________________________ 2

1.4 ขอบเขตของโครงงาน _________________________________________________________ 2

1.5 ขนตอนและวธการดาเนนงาน ___________________________________________________ 3

บทท 2 ทฤษฏทเกยวของ _________________________________________________________________ 5

2.1 Functional electrical stimulation of walking: Function, exercise and rehabilitation_________ 5 2.1.1 Historically used___________________________________________________ 5 2.1.2 Terminology______________________________________________________ 5 2.1.3 Why use FES?_____________________________________________________ 7 2.1.4 Lower Extremity Applications________________________________________13 2.1.5 UE research results_________________________________________________16

2.2 Functional electrical stimulation of walking: Function, exercise and rehabilitation________ 20 2.2.1 Abstract_________________________________________________________ 20 2.2.2 Anatomy_________________________________________________________21 2.2.3 Peroneal neuropathies______________________________________________ 26 2.2.4 Examination of the patient with foot drop ______________________________31

-C-

2.2.5 Differential diagnosis of foot drop_____________________________________32 2.2.6 Investigations_____________________________________________________33

2.2.7 Management of the patient with foot drop_______________________________38 2.3 Functional electrical stimulation of walking: Function, exercise and rehabilitation________ 41

2.3.1 Abstract_________________________________________________________ 41 2.3.2 Introduction______________________________________________________ 41 2.3.3 Brief history of FES-assisted walking_________________________________ 42 2.3.4 FES-assisted paraplegic gait_________________________________________ 43 2.3.5 Fully implanted systems____________________________________________ 44 2.3.6 Foot drop stimulators_______________________________________________44 2.3.7 Hybrid orthotic systems_____________________________________________45 2.3.8 Therapeutic applications____________________________________________ 46 2.3.9 Limitations_______________________________________________________47 2.3.10 Conclusions______________________________________________________ 47

2.4 Functional electrical stimulation of walking: Function, exercise and rehabilitation________ 48 2.4.1 Watch the Step-by-Step Video________________________________________48 2.4.2 Step-by-Step Guide________________________________________________ 48 2.4.3 Download________________________________________________________62 2.4.4 Related Examples__________________________________________________62

บทท 3 แนวทางการพฒนา_______________________________________________________________ 63

3.1 การพฒนาซอรฟแวร__________________________________________________________63

3.1.1 ซอรฟแวรทเขยนใน MCU ตวท 1________________________________________63

3.1.2 ซอรฟแวรทเขยนใน MCU ตวท 2________________________________________68

3.2 การพฒนาฮารตแวร__________________________________________________________ 74

3.2.1 ศกษาและเลอกอปกรณทจะใชในการทา__________________________________ 74

3.2.2 ออกแบบ PCB และจดอปกรณตางๆ_____________________________________ 80

-D-

บทท 4 การทดลองและผลการทดลอง_______________________________________________________82

4.1 การทดลองในโหมด Calculation_________________________________________________82

4.2 การทดลองในโหมด Normal____________________________________________________82

4.3 การทดลองในโหมด Manual switch______________________________________________ 83

4.4 การทดลองในโหมด Tilt sensor_________________________________________________ 84

บทท 5 บทวจารณและสรปผล_____________________________________________________________86

5.1 ปญหาและการแกไข__________________________________________________________ 86

5.2 แนวทางการพฒนาตอ_________________________________________________________ 86

5.3 สรปผลและวจารณ____________________________________________________________86

บรรณานกรม__________________________________________________________________________87

-E-

สารบญภาพ

ภาพท หนา

ภาพท 2.1 The sacral plexus and the origin of the sciatic nerve___________________________________21

ภาพท 2.2 Posterior view of the course and branches of the right sciatic nerve_______________________22

ภาพท 2.3 Anterolateral view of the right leg_________________________________________________23

ภาพท 2.4 Patient with a right foot drop and sensory abnormality_________________________________23

ภาพท 2.5 Gesture of foot drop patient 1____________________________________________________ 29

ภาพท 2.6 Gesture of foot drop patient 2____________________________________________________ 29

ภาพท 2.7 Gesture of foot drop patient 3____________________________________________________ 30

ภาพท 2.8 Comparison between normal and abnormal peroneal nerve_____________________________ 34

ภาพท 2.9 Proton density-weighted MR axial image of the leg below the knee______________________ 35

ภาพท 2.10 Foot drop___________________________________________________________________ 40

ภาพท 2.11 Overall system model__________________________________________________________49

ภาพท 2.12 Configure the ADC Block______________________________________________________50

ภาพท 2.13 Configure Target HID Send Block_______________________________________________ 51

ภาพท 2.14 Connected model_____________________________________________________________52

ภาพท 2.15 Updated Model_______________________________________________________________53

ภาพท 2.16 Build target complete__________________________________________________________54

ภาพท 2.17 Host PC system model_________________________________________________________55

ภาพท 2.18 Configure Host HID Receive Block______________________________________________ 56

-F-

ภาพท 2.19 Configure Data Type Conversion Block___________________________________________ 57

ภาพท 2.20 Completed host PC model______________________________________________________ 58

ภาพท 2.21 Updated diagram before simulation_______________________________________________59

ภาพท 2.22 Connect FiO Std potentiometer output to pin C3_____________________________________60

ภาพท 2.23 ADC HID Demo system test____________________________________________________ 61

ภาพท 3.1 แสดงภาพรวมของ MCU 1 บนโปรแกรม Simulink_________________________________ 64

ภาพท 3.2 แสดงการทางานของ State Flow ใน MCU 1________________________________________ 65

ภาพท 3.3 แสดงการทางานใน Normal mode________________________________________________ 66

ภาพท 3.4 แสดงการทางานในโหมด Manual switch__________________________________________67

ภาพท 3.5 แสดงการทางานในโหมด Tilt sensor_____________________________________________ 68

ภาพท 3.6 แสดงภาพรวมของ MCU 2 บนโปรแกรม Simulink___________________________________ 69

ภาพท 3.7 แสดงการทางานของ State Flow ใน MCU 2________________________________________ 71

ภาพท 3.8 แสดงการทางานในโหมด Calculation____________________________________________ 72

ภาพท 3.9 แสดงการทางานในโหมด Normal_______________________________________________ 72

ภาพท 3.10 แสดงการทางานในโหมด Manual switch_________________________________________73

ภาพท 3.11 แสดงการทางานในโหมด Tilt sensor____________________________________________73

ภาพท 3.12 แสดง LCD ทใช_____________________________________________________________ 74

ภาพท 3.13 แสดง Schematic ของ LCD_____________________________________________________74

ภาพท 3.14 แสดงบอรด Fio std ทใช_______________________________________________________ 75

ภาพท 3.15 แสดงรายละเอยด PIN และ JP บนบอรด__________________________________________ 75

-G-

ภาพท 3.16 แสดงภาพ Accelerometer Sensor ทใช_____________________________________________76

ภาพท 3.17 แสดงภาพของตวหมนปรบคาพารามเตอรตางๆทใช__________________________________ 77

ภาพท 3.18 แสดงรป Switch เปลยนโหมดทใช________________________________________________77

ภาพท 3.19 แสดง LED ทแสดงการทางานในโหมดตางๆ_______________________________________ 78

ภาพท 3.20 แสดงอปกรณปรบเปลยนแอมพลจด______________________________________________ 78

ภาพท 3.21 แสดงรป Electrode ทใช________________________________________________________79

ภาพท 3.22 แสดง Power supply ทใช_______________________________________________________ 79

ภาพท 3.23 แสดง PCB ทออกแบบเรยบรอยแลว______________________________________________80

ภาพท 3.24 แสดงภาพรวมของอปกรณทเสรจสมบรณ ( Side view )_______________________________81

ภาพท 3.25 แสดงภาพรวมของอปกรณทเสรจสมบรณ ( Top view )_______________________________81

ภาพท 4.1 แสดงการทางานในโหมด Calculation______________________________________________82

ภาพท 4.2 แสดงการทางานในโหมด Normal_________________________________________________83

ภาพท 4.3 แสดงสญญาณทไดเมอตงคาไวคาหนง______________________________________________83

ภาพท 4.4 แสดงการทางานในโหมด Manual switch___________________________________________ 84

ภาพท 4.5 แสดงการทางานในโหมด Tilt sensor_______________________________________________85

ภาพท 4.6 แสดงการวดความกวางและความถสญญาณ__________________________________________85

-H-

บทท 1

บทนา

1.1 ทมาและภาพรวมของโครงการ

หลงจากทไดรวมงานกบทางบรษทเอมมาจน ทางดร.กฤษฏากไดใหพวกผมเลอกทาระบบประยกต

มา หนงอยางทใชโปรแกรม Simulink แทนการใชภาษาซ ซงพวกผมไดเลอกทจะสรางเครองกระตน

กลามเนอสาหรบผปวยโรคกลามเนอออนแรง

กอนอนผมขอแนะนาใหรจกโรคกลามเนอออนแรงกอนวาเปนอยางไรเพราะวาหลายๆคนยงไมรจก

โรคน โรคกลามเนอออนแรงหรอ ALS (Amyotrophic Lateral Sclerosis) ไมใชโรคของกลามเนอโดยตรง แต

เปนโรคทเกดจากความผดปกตของเซลลประสาทนาคาสง แลวสงผลทาใหกลามเนอออนแรงเนองจากขาด

เซลลประสาทนาคาสงมาควบคม ซงเซลลเหลานมอยในไขสนหลงและสมอง โดยทเซลลประสาทนาคาสง

เหลานคอยๆ เกดการเสอมและตายไปในทสด และเนองจาก โรคกลามเนอออนแรง เอแอลเอสเปนโรคทเกด

จากความผดปกตของ เซลลประสาทนาคาสง จงมชอเรยกอกอยางหนงวา "โรคของเซลลประสาทนาคาสง

(motor neuron disease; MND) หรอ โรคเซลลประสาทนาคาสงเสอม" ในประเทศสหรฐอเมรกาจะรจกโรคน

ในชอของโรค ล-เก-รก (Lou Gehrig Disease) ซงตงชอโรคตามชอนกเบสบอลทมชอเสยงทเปนโรคนในป

ค.ศ. 1930

ใน ปจจบนยงไมทราบแนชดวาเหตใดเซลลประสาทนาคาสงจงเกดการเสอมโดย สมมตฐานเชอวา

โรคกลามเนอออนแรงเอแอลเอส เกดจากหลายเหตปจจยกอใหเกดโรครวมกน ไดแก การมปจจยบางอยางทาง

พนธกรรมซงยงไมทราบแนชดททาใหเซลลประสาทนา คาสงมโอกาสเสอมไดงายกวาบคคลอน มปจจยทาง

สงแวดลอมทมการปนเปอนของสารพษ เชน ยาฆาแมลง สารโลหะหนก รงสหรอการตดเชอไวรสบางชนดมา

ชวยกระตนสงเสรมใหเซลลประสาทนาคา สงเกดการทางานผดปกต รวมกบอายทสงขนตามกาลเวลาทาให

เกดการเสอมสลายของเซลล อนเนองมาจากแบตเตอรทคอยสรางพลงงานใหกบเซลลทเรยกวา ไมโตคอนเด

รย (mitochondria) มความผดปกต แตสมมตฐานเหลานยงไมไดรบการพสจนทแนชด

สาหรบอาการนน เรมตน ผปวยจะมอาการกลามเนอออนแรงของมอ แขน ขา หรอเทาขางใดขางหนง

กอน เชน ยกแขนไมขนเหนอศรษะ กามอถอของไมได ขอมอหรอขอเทาตก เดนแลวหกลมบอยหรอสะดด

-1-

บอย ขนบนไดลาบาก ลกนงลาบาก เปนตน อาการกลามเนอออนแรงจะคอยๆ เปนมากขนจนลามไปทง 2 ขาง

ผปวยบางรายอาจมอาการกลามเนอออนแรงของแขนหรอขาทงสองขางตงแต ตน นอกจากอาการกลามเนอ

ออนแรงแลวยงพบวามกลามเนอลบรวมกบกลามเนอ เตนรวมดวย ผปวยบางรายอาจมาพบแพทยครงแรกดวย

มอลบหรอขาลบ พดไมชด พดเหมอนลนแขง ลนลบ เวลากลนนาหรออาหารแลว จะสาลก

สาหรบผทมโอกาสเปนโรคนนน ขอมล ในประเทศสหราชอาณาจกรพบประชากรทกๆ 100,000 คน

เปนโรคเอแอลเอส ประมาณ 2 คนตอป อายเฉลยทเกดขนของโรคอยระหวาง 60-65 ป ดงนนโอกาสทจะพบ

โรคเอแอลเอส ในคนอายมากจงมมากกวาในคนอายนอย โดยทวไปแลวมกพบโรคเอแอลเอส ไดบอย

ประมาณ 1.5 เทาของเพศหญง และประมาณรอยละ 90 ของผปวยเอแอลเอส จะไมไดเกดจากสาเหตทแนชด

ทางพนธกรรม ดงนนผทไมมประวตครอบครวทชดเจน จงมโอกาสเสยงนอยมากทจะเกดโรคในรนลกรน

หลาน นอกจากนยงไมมหลกฐานพสจนไดชดเจนวานกกฬามโอกาสเสยงตอการ เกดโรคนไดมากกวาอาชพ

อนๆ

จากขอมลจะเหนไดชดวาในปจจบนมผทปวยเปนโรคกลามเนอออนแรงเพมมากขน และในตลาด

เครองกระตนกลามเนอสาหรบผปวยทมตามโรงพยาบาลนนมราคาสง ดงนนพวกผมจงเหนความสาคญของ

การทาเครองกระตนกลามเนอขนมาในราคาทเหมาะสมกบคณภาพทผใชจะไดรบ

วตถประสงคโครงการ 1.2

• สามารถพฒนาการใชโปรแกรม Simulink แทนการใชภาษาซในการสรางระบบประยกตได

• เพอสรางเครองกระตนกลามเนอสาหรบผปวยโรคกลามเนอออนแรง

• สามารถจดทาเอกสารเพอรวบรวมขอมล และขนตอนการทางานจากการพฒนาโครงการเพอ

เปนประโยชนแกผทสนใจตอไป

1.3 ประโยชนทคาดวาจะไดรบ

• ความรทางดาน Simulink ในการประยกตใชเขากบระบบสมองกลฝงตว

• การทา Printed circuit board (PCB) และความรดานฮารตแวรตางๆ

• ความรทางดานโรคกลามเนอออนแรง เพอทจะไดนาไปใชสรางเครองกระตนกลามเนอ

1.4 ขอบเขตของโครงงาน

• สามารถสรางเครองกระตนกลามเนอได

-2-

• สามารถใหผใชปรบความแรงของสญญาณได

• สามารถแสดงขอมลตางๆบนหนาจอ Light-emitting diode (LED) ได

• สามารถใชเซนเซอรในการรบคาเพอใหเครองกระตนอตโนมตได

• สามารถแสดงไฟการทางานวาอยในโหมดไหน กระตนอยหรอไมกระตน ได

1.5 ขนตอนและวธการดาเนนงาน

ขนตอนการทางานของพวกผมจะเปนเปน 4 ชวงดวยกน ดงน

- ขนแรก ศกษาและคนควาหาขอมล

• หาขอมลเกยวกบโรคกลามเนอออนแรง

• ศกษาตลาดวามผผลตเครองออกมาเปนแบบใดบาง

• วางแผนและออกแบบเครองทจะสรางและงานทจะทา

• แบงงานทจะทาออกเปนสวนๆ

- ขนทสอง การพฒนาซอรฟแวร

• กาหนดเปาหมายของผลลพธทตองการ

• ศกษาการใช Simulink ทางดานระบบสมองกลฝงตว

• ทดลองโปรแกรมในรปแบบตางๆจนไดวธทเหมาะสม

• ปรบปรงและแกไขโปรแกรมจนไดผลลพธทพอใจ

- ขนทสาม การพฒนาฮารตแวร

• วางแผนการเชอมสวนของฮารตแวรและซอรฟแวร

• ออกแบบวงจรและการวางอปกรณตางๆทจาเปนตองใช

• ทดสอบการทางานและผลทได

• ปรบปรงแกไข

- ขนทส การจดทาเอกสาร

• จดทาแผนการทางาน

• จดทาเอกสารและการนาเสนอผลงาน

-3-

แผนการดาเนนงาน

 

-4-

บทท 2

ทฤษฏทเกยวของ

2.1 Functional Electrical Stimulation

“To stim or not to stim?”

2.1.1 Historically used

Functional Electrical Stimulation (FES) has been used as a treatment modality for gait and chronic

stroke since the 1960’s. Lieberson was one of the first to investigate the use of FES for foot drop. There has

been renewed interest recently in electrical stimulation with new technologies and software largely driven

by research in spinal cord injury.

2.1.2 Terminology

Neuromuscular Electrical Stimulation or NMES

The electrical stimulation of the lower motor neuron, to activate paralyzed or weakened muscles.

Functional Electrical Stimulation or FES

FES involves the use of NMES to activate muscles in a specific order and degree to complete a

functional task. Functional Electrical Stimulation is a therapeutic modality that can be used effectively in

neurological conditions with intact lower motor neurons, healthy neuromuscular junctions and muscle

-5-

tissue. This narrows its application to conditions such as spinal cord injury, stroke, head injuries, cerebral

palsy and multiple sclerosis.

This is an open loop system that uses vision and proprioception for sensory feedback.

Neuroprosthesis

This is an apparatus or system that provides FES to enhance functional activity. This is a closed

loop system that provides “real-time modification of the stimulation pattern based on sensory feedback.” It

can respond more quickly and adapt to changing demands due to the increased sources of feedback.

Physiology

FES involves applying an electrical current, with electrodes on or close to the innervating nerve

fibres, to elicit an action potential, producing a muscle contraction that can be modified by changing

stimulus parameters, in order to help restore functional movement.

An electrical field produced by the stimulating electrode depolarizes the cell membranes of

neurons. The nerve is stimulated as it has a lower firing threshold than muscle. When depolarization

reaches threshold, an action potential is produced by the influx of sodium from extracellular to intracellular

space. The action potential travels distally to the neuromuscular junction and causes contraction of muscle

fibres. The large motor units (type II) are activated first as they fire at a lower threshold. These are the

motor units used for speed, power and tend to fatigue more quickly. This is the reverse of the Henneman

size principle of recruitment order in voluntary muscle contraction. With disuse atrophy there is also a

conversion of type I slow twitch fatigue resistant fibres to type II. There has been some research that has

found chronic stimulation can reverse the fibre type conversion.

-6-

2.1.3 Why use FES?

In an article by J Daly and RL Ruff, the authors discuss functional interventions based on brain

plasticity and motor learning principles. They identify the goal of rehab is to “design and test interventions

than result in impairment gains sufficiently robust to be reflected in functional activity and further in life

role participation”. For CNS activity dependent plasticity, the critical principles of motor learning are

• Close to normal movements

• Muscle activation during practice of movement

• Focused attention

• Repetition of desired movement

• Training specificity

Research has shown us that recovery is aided by motor experience. Repetition of movements has

also been identified and a key in motor relearning.

FES may facilitate motor recovery with muscle and joint afferent feedback with repetitive

movement.

Peripheral stimuli can influence reorganization in the brain.

The afferent and efferent input from movement facilitated by FES can play an important role as a

reminder on “how to perform movement properly.”

Daly and Ruff support FES as fulfilling the motor learning principles for gait. According to the

literature, FES has been used to treat weakness, incoordination of movement, abnormal tone, gait problems.

-7-

It can closely reproduce the motor components of gait. It does provide the “practice of close-to-normal

movement and repetition of that practice.”

Neurophysiology

Research findings:

Increases in muscle oxidative capacity

Increased norepinephrine levels in plasma

Increased size and number of type IIa fibers

Set up and Parameters

In order to have a flow of current, 2 electrodes are required. The electrode arrangement can be

monopolar or bipolar.

Monopolar – the active electrode (tends to be a smaller electrode) is positioned by the peripheral

nerve, the indifferent or return electrode is positioned over less excitable tissue (tendon or fascia). In a

multichannel monopolar system, there is only one indifferent electrode and several active electrodes.

Bipolar –the active electrode is also positioned by the nerve to be stimulated and the indifferent

electrode is placed close to the active electrode. In a multichannel bipolar system, for each active there is an

indifferent electrode. This may provide a more selective area of muscle activation.

-8-

Provision of electrical stimulation is in a waveform of electrical current pulses. The strength of the

muscle contraction is controlled by the pulse frequency, amplitude and duration of the current pulses.

Pulse frequency

• Need a high enough frequency, fusion frequency, to create a smooth contraction. Too low and you

produce a series of twitches.

• Temporal summation is the cumulative effect of stimuli repeated in a short time period

• The higher the stimulus frequency, the stronger the contraction but the muscle also fatigues more

quickly.

• The minimum stimulus frequency rates are usually 12-15 Hz to achieve fused muscle response.

• Ideal stimulation frequency for upper extremity 12-16 Hz and 18-25 for lower extremity muscle

stimulation.

Pulse amplitude and duration

• Spatial summation is the effect of increasing the number of activated motor units, to increase the

strength of contraction.

• Increasing the pulse amplitude and or pulse duration increases the number of axons and motor units

activated due to the effect of a larger charge and resulting electrical field being produced. This

increases the area of activation and increases the strength of contraction.

• Amplitude beyond motor threshold also excites small diameter unmyelinated C fibers which elicit

pain

• Pulse duration of 200-400, 300 us appears to be more comfortable

Waveform

• Monophasic – repetitive unidirectional pulse, which is primarily cathodic or negative phase

• Biphasic – repetitive pulse with a cathodic then anodic (positive) phase

-9-

• The first phase brings about the action potential

• The second phase balances out the charge to prevent tissue damage which is key when using

implanted electrodes

Stimulators

These control the current or voltage

• With voltage regulated stimulators the amount of delivered current is dependent on electrode

interface impedance

With surface electrodes, as the electrodes dry or lose contact with skin, the

impedance increases and the current flow decreases thereby reducing the risk of

burns

Due to the electrode-skin interface and variable impedance, the motor response is

less consistent

Used with surface or transcutaneous electrodes

• Current is better regulated and not affected by impedance with current regulated stimulators

Used with implanted electrodes

More consistent motor response

Stimulation delivery

• Surface – electrodes are placed over nerves or motor points

Noninvasive, easily applied, somewhat inexpensive, easily used in clinic settings

-10-

Repeated consistently accurate electrode placement can be difficult

Can be painful with sensitive skin

Challenging to attain deep muscle activation or specific muscle contraction

Aesthetics and management of a system with multiple components

Electrodes can shift or fall off while limb is moving

• Percutaneous – use electrodes that are implanted into isolated muscles with hypodermic needle

Provide activation for deep muscle, repetitive accurate muscle contractions

Require lower stimulation currents

Less painful

Exit site of electrode lead must be kept clean

Risk of breakage or infection

Last for up to 3 months

Used for trial before implantable systems surgery

• Implantable systems – electrodes and stimulator are implanted

An external control unit provides power and instruction through a radio-frequency

telemetry link

Electrodes can be implanted on or in the muscle, beside or around a nerve

Good for wide, thin or superficial muscle

Increased specificity of muscle stimulation

No surface wires, a small antenna is taped over the stimulator site

Risk of tissue growth affecting the nerve

Long term use

Contraindications

Pacemaker

Peripheral vascular disease if possibility of causing thrombi to loosen

-11-

Hypertension or hypotension can affect autonomic responses

Areas of excessive adipose tissue increase impedance

Neoplastic tissue

Active infection

Devitalized skin due to x-ray therapy

Cognitive issues affecting ability to provide feedback

Not over carotid sinus

Not over thoracic region

Not over phrenic nerve

Not over trunk if pregnant

Example of parameter settings.

Modulation: Pulse or burst

Amplitude: maximum tolerated

Pulse Duration: 100-300 us

Frequency: 20-100

Duty cycle: on:off 2:10 or 5:15

Number of contractions: 10-20 at maximum intensity, 15minutes session 2-3x day

Frequency : 3-7x week

de Kroon et al reviewed the relationship between stimulation parameters and clinical outcome in

studies using electrical stimulation (ES) to improve motor control of upper extremity in stroke. Motor

stimulation can be applied by neuromuscular electrical stimulation with pre-programmed system with no

active client involvement, EMG triggered electrical stimulation or positional feedback stimulation training

both of which involve voluntary muscle contraction to trigger the ES. In review of 19 studies that fit the

criteria, there was no relationship detected between the stimulation parameters (frequency, amplitude, pulse

-12-

duration), duration of stimulation, subject characteristics and clinical outcome. There was a suggestion that

a triggered stimulation may be more effective than the non-triggered stimulation in improving motor

control.

Research does suggest that parameters are important to the muscle contraction. But muscle

contraction is the desired outcome and parameters are adjusted to achieve this. Often FES units have

preprogrammed protocols. There is some research looking at trying to find parameters that maximize

contraction and minimize fatigue.

The triggered stimulus introduces the aspect of the cognitive component. Animal research studies

have shown behavioural experiences which are meaningful, and require a cognitive intent, tend to have task

specific cortical reorganization with development of motor skills.

In a study looking at repeated stimulation of peroneal nerve, Khaslavskaia and Sinkjaer concluded

that “cortical excitability modulated by peripheral nerve stimulation is focally and task specifically affected

by voluntary cortical drive.”

The importance of being actively engaged in initiating or producing the movement may be a

significant component in promoting recovery.

2.1.4 Lower Extremity Applications

In the Evidence Based Review of Stroke Rehabilitation 10th edition ( EBRSR) review of FES in the

lower extremity, they state that “There is strong level 1A evidence that FES and gait training results in

improvements in hemiplegic gait”.

Initial use of FES was for electrical stimulation of peroneal nerve to treat weakness of ankle

dorsiflexion or foot drop. There are systems available that consist of an heel sensor, stimulator and

electrodes. The electrodes are place over common peroneal nerve cathode near head of fibula and anode

-13-

over tibialis anterior. The heel sensor detects the pressure and tilt of the leg to turn on tibialis anterior at toe

off and turn off tibialis anterior at heel strike.

Results of studies of lower limb electrical stimulation

Improved walking ability

Increased isometric contractions of dorsiflexors and plantarflexors

Decreased cocontraction

With EMG triggering

Increased mobility

Increased voluntary EMG activity

With a neuroprosthesis system, adding stimulation of knee and hip muscles

Improved gait performance

Motor learning progressed to gait training without neuroprosthesis

Use of FES combined with physiotherapy or biofeedback produces superior results in ambulation

scores than any individual treatment alone.

Burridge et al found increased walking speed with stimulation of peroneal nerve.

This is not a common treatment for foot drop in North America. Some of the issues are accurate

electrode placement, medial and lateral ankle control, requires custom AFO.

There is evidence of motor relearning with peroneal nerve stimulation. The problems are accurate

surface electrode placement, costs of invasive procedures for the implanted systems, complications of

-14-

dealing with medial and lateral control as well as dorsiflexion. There is a distinct lack of large randomized

control studies to look at the effectiveness of peroneal nerve stimulation for dorsiflexion.

Functional Electrical Therapy (FET) is patterned multi-channel electrical stimulation acting as a

neural prosthesis activating several muscles during gait. It assists with walking when there is little or no

muscle activity in the leg. This can prevent disuse atrophy, discourage compensatory gait patterns, maintain

some sensory motor input. Preliminary findings for a recent trial with hemiplegic patients, Popovic, show

notably better recovery of walking with FET than with only conventional therapy. Popovic proposes that in

the acute stage, the combined effect of therapy generated and spontaneous recovery can accelerate

functional recovery. FET may be facilitating cortical reorganization in the acute stages of recovery.

Results

Changes in corticospinal excitability can be achieved with repetitive stimulation of the common

peroneal nerve. This may be a result of activating the motor and sensory fibers and changes in muscle fibre

properties.

Benefits are increase strength, decrease spasticity, increased muscle extensibility

Cortical reorganization, cortical connectivity may be modified with high frequency sensory

stimulation.

Stimulation of ankle PF improves ground clearance, decreased extensor tone and decreased swing

time.

These of some of the ankle dorsiflexion systems

Footlifter (Elmetec A/S Denmark) – surface system with one channel and a heel switch

Walkaide ( University of Alberta) – tilt sensor and cuff surface system worn below the knee

-15-

Odstock Footdop Stimulator (Salisbury UK) - single channel footswitch–triggered surface

stimulator system. It has been used in a randomized control study and results showed increased speed of

ambulation and decrease in effort of walking.

The National Clinical FES centre in England uses the “Odstock Dropped Foot Stimulator”.

2.1.5 UE research results

EBRSR review of FES in treatment of the upper extremity,” There is strong (Level 1A) evidence

that FES treatment improves upper extremity function in acute stroke”.

“There is conflicting (level 4) it evidence that functional electrical stimulation reduces pain

improves function and reduces subluxation following stoke.”

There are limited randomized studies, with methodology variability but the results do indicate that

there is some positive effect on upper extremity motor learning. With more significant results in the acute

stage and those with milder impairment. These were studies looking at wrist extension.

In studies with biofeedback or EMG triggering NMES, there were gains made in motor recovery

with some improvement in activity.

A Cochrane review of electrical stimulation for preventing and treating post stroke shoulder pain

found 4 randomized control studies (170 subjects). There was no significant change in pain intensity or

-16-

incidence, there was a significant effect in improving passive pain-free range of lateral rotation of the

humerus with a decrease in subluxation. There was no significant effect on motor recovery.

Some of the difficulty with using FES for the hemiplegic shoulder are:

Cutaneous pain receptors are stimulated which affects tolerance and compliance

Difficult to activate deep muscles

Difficult to control the level of contraction

Accurate electrode placement and finding the stimulation parameters that are tolerated and

achieve the desired affect requires skill and experience

There has been a trial done with a percutaneous system with a reduction in pain in chronic stroke

survivors.

Some of the technology used for upper extremity stimulation systems for the spinal cord

population, neuroprostheses, have had limited application to stroke population. The movement is limited to

a few functional activities namely opening and closing the hand. If tone is an issue, the increased effort to

do the task often increases the tone, affecting the quality and outcome of the intended movement. In some

trials with a neuroprosthesis there have been significant improvements reported. According to Sheffler

there is no clinically viable hand neuroprosthesis system currently available to meet the needs of the

hemiplegic arm and hand, more research is needed. A feasible system would require the ability to facilitate

bilateral tasks, permit distal and proximal control, be a compact size to ensure ambulation is not

encumbered, stimulate weak and inhibit overactive muscle and not compromise intact arm.

These are some of the upper extremity neuroprostheses that have been developed.

Handmaster ( NESS Ltd, Israel, recently available BioNESS USA)– surface system used for finger

thumb muscle activation. It consists 5 electrodes built into a wrist hand orthotic used to stimulate hand

opening and closing. The design provides wrist stabilization at a 10 20 angle of extension.

-17-

FESMate (NEC Medical Systems Japan) – 30 electrode percutaneous system. Activates hand and

upper extremity movements.

Freehand (Case Western Reserve University Cleveland VA Medical Center) – implanted system for

lateral and palmar grasp with C5-C6 tetraplegia.

Results

Improved hand and upper limb function in ADLs

Decreased spasticity

Decreased shoulder pain

Better outcomes with FES combined with functional task component practice

Disadvantages to use of FES

Muscle fatigue – increases with increased frequency and intensity of stimulation

Pain

skin irritation

unrealistic expectations

multichannel surface systems can be cumbersome to apply and wear

accurate electrode placement

Conclusions

Although there have been positive results with use of FES, to generalize and present a protocol for

FES as a treatment is difficult. The problem lies with variability in methodology, limited random control

studies, lack of standardized treatment protocols, variability in population utilized ( acute vs chronic).

-18-

There is a definite need for large, multicentre, randomized clinical trials with defined population

and evaluation of immediate and long term outcomes.

It is difficult to determine exact parameters for electrical stimulation due to many confounding

variables. There is further research required to examine and determine optimal dose and parameters.

Techniques and systems need to be improved to encourage compliance.

Lack of studies comparing FES with EMG triggered FES and neuroprostheses.

Research has shown:

FES can be used for the hemiparetic population that demonstrates limited residual movement.

There does appear to be some limited evidence to support use of FES for shoulder pain and to

decrease subluxation.

Better outcomes with task specific activities than with generalized function.

When using FES, repetition, novelty, active contribution and function are key points for achieving

motor relearning.

Future of FES

There is a definite lack of large scale research to definitively support or dismiss the use of FES in

treatment of hemiplegic limbs in stroke. There appears to be a push toward increased use of technology

with robot arms and neuroprostheses as treatment adjuncts. The cost of this technology may prevent

immediate access for clients in our health care system. Regardless, the emphasis is still on tapping into

motor learning and neuroplasticity using techniques of repetition and task specific training to improve

functional outcomes. Research does provide some support for utilizing FES in promoting motor relearning

and with further refined research there will likely be the development of more definitive protocols for

treatment.

-19-

2.2 Foot drop: where, why and what to do? 1. John D Stewart

+Author Affiliations 1. Consultant Neurologist, Lions Gate Hospital, North Vancouver, British Columbia, Canada

1. Dr J D Stewart, 145 East 13th Street, #204, North Vancouver, BC V7L 2L4, Canada;john.stewart@telus.net

2.2.1 Abstract

Foot d p is a comro mon and distressing problem that can lead to falls and injury. Although the most frequent cause is a (common) peroneal neuropathy at the neck of the fibula, other causes include anterior horn cell disease, lumbar plexopathies, L5 radiculopathy and partial sciatic neuropathy. And even when the nerve lesion is clearly at the fibular neck there are a variety of causes that may not be immediately obvious; habitual leg crossing may well be the most frequent cause and most patients improve when they stop this habit. A meticulous neurological evaluation goes a long way to ascertain the site of the lesion. Nerve conduction and electromyographic studies are useful adjuncts in localising the site of injury, establishing the degree of damage and predicting the degree of recovery. Imaging is important in establishing the cause of foot drop be it at the level of the spine, along the course of the sciatic nerve or in the popliteal fossa; ultrasonography, CT and MR imaging are all useful. For patients with a severe foot drop of any cause, an ankle foot orthosis is a helpful device that enables them to walk better and more safely.

The colloquial and medical term “foot drop” admirably describes weakness of the dorsiflexor muscles of the foot. It is to be distinguished from flail foot, in which all the muscles below the knee are

the

e se

• uropathy, or could there be a more proximal lesion?

•

•

•

-20-

affected—the plantar flexors as well as the dorsiflexors. For neurologists, the keen-eyed lot that we are, most familiar sighting of a foot drop is likely to be that of the late John Thaw who played the part of Inspector Morse in numerous BBC films; we do not know what the cause was. Although a lesion of th(common) peroneal nerve is in general the most likely cause of foot drop, the wary neurologist should pothe following questions:

is this really a peroneal ne

if it is a peroneal neuropathy, what is the cause?

what investigations will sort this out?

what can be done to help the patient?

2.2.2 ANATOMY The tibialis anterio

This is derived from anterir, the main dorsiflexor muscle of the foot, is innervated by the peroneal nerve.

or horn cells in the lower spinal cord. Their axons travel in the L4 and L5 spinal nerve roots (“roots”), they then join to form the lumbosacral trunk that connects these lumbar plexus structures to the sacral plexus (fig 1). These nerve fibres then enter the lateral trunk of the sciatic nerve which becomes the peroneal nerve when the sciatic divides just above the knee (fig 2). The lateral trunkgives off only one branch in the thigh—that to the short head of the biceps femoris muscle. All of the othhamstring muscles are innervated by the medial trunk of the sciatic nerve; that trunk becomes the tibial nerve.

er

Figure 2.1The sacral plexus and the origin of the sciatic nerve. (Reproduced with permission from Stewart JD. Focal peripheral neuropathies. Third edition. Philadelphia: Lippincott Williams & Wilkins, 2000.)

-21-

Figure 2.2Posterior view of the course and branches of the right sciatic nerve. (Reproduced with permission from Stewart JD. Focal peripheral neuropathies. Third edition. Philadelphia: Lippincott Williams & Wilkins, 2000.)

The peroneal nerve passes laterally through the popliteal fossa and winds around the head and neck of the fibula (fig 3). It is closely applied to the periosteum of that bone for about 6 cm and, for most of this distance, it is covered only by skin and subcutaneous tissue. It then pierces the peroneus longus muscle to reach the anterior compartment of the lower leg. At that point, the fibres of the muscle form a tendinous arch over the nerve, and this has been termed the fibular tunnel (fig 3). The nerve supplies the tibialis anterior, the extensors of the toes, and the foot everter (peroneal) muscles. It also supplies the skin over the anterolateral aspect of the lower leg from about midway between the knee and the ankle, and most of the dorsal aspect of the foot and toes. This extensive distribution of sensory loss is seen when the nerve is lacerated at the knee, However, when it is compressed, the sensory loss is much more restricted—usually just to the dorsum of the foot and toes (fig 4). In some patients there may be no sensory signs or symptoms at all, presumably because of sparing of relevant nerve fascicles within the damaged portion of the nerve.

-22-

Figure2.3Anterolateral view of the right leg showing the course, clinically relevant anatomical relations, and major branches of the common peroneal nerve. (Reproduced with permission from Stewart JD. Focal peripheral neuropathies. Third edition. Philadelphia: Lippincott Williams & Wilkins, 2000.)

Figure 2.4Patient with a right foot drop and sensory abnormality in the territory of the distal superficial peroneal nerve (dorsum of the foot) and the deep peroneal nerve (web space between first and second toes

-23-

and a small area in the adjacent dorsum of the foot). This is the same patient as in figure 5. Informed consent was obtained for publication of this figure.

Therefore, an anatomy-based differential diagnosis of foot drop includes lesions or disorders affecting anterior horn cells, L4 or L5 roots, lumbosacral plexus, sciatic nerve and peroneal nerve. In practice, in many cases, the cause of the foot drop is clear—that is, a patient who has fallen and struck the lateral knee, or a patient with acute low back pain and classic lumbar radiculopathy symptoms. In many other patients in whom the diagnosis is less obvious, a peroneal neuropathy remains the likely diagnosis, but there are several pitfalls waiting to confound even the experienced clinician.

TABLE Causes of peroneal neuropathy

External compression

During anaesthesia, coma, sleep, bed rest

Plaster casts, braces

Habitual leg crossing

Sitting cross-legged

Prolonged squatting, kneeling

Direct trauma

Blunt injuries, lacerations

Fractures of the fibula

Adduction injuries and dislocations of the knee

Surgery and arthroscopy in popliteal fossa and knee

-24-

Traction injuries

Acute ankle injuries

Masses

Ganglia, Baker’s cysts, callus, fibular tumours, osteomas, haematomas

Tumours

Nerve sheath tumours

Nerve sheath ganglia

Lipomas

Entrapment

In the fibular tunnel

Anterior (tibial) compartment syndrome

Vascular

Vasculitis, local vascular disease

Diabetes mellitus: susceptibility to compression, ischaemic damage

Leprosy

Idiopathic

-25-

2.2.3 PERONEAL NEUROPATHIES The many causes of this common focal neuropathy are listed in the table.

• Acute trauma is a frequent cause and includes direct blows and lacerations, severe adduction injuries and dislocations of the knee, fractures of the head or neck of the fibula, and bullet wounds. The common peroneal nerve can also be inadvertently injured during knee operations, including total knee replacement and arthroscopic surgery. A specific type of injury results from acute plantar flexion and inversion injuries at the ankle, usually severe sprains or fractures of the distal tibia and fibula; the foot drop is usually immediate, but may not appear for several days. It seems that extreme and sudden ankle inversion exerts enough traction on the peroneal nerve in the popliteal fossa to tear the vasa nervorum where they enter the nerve sheath. • External pressure is probably the most frequent cause of peroneal neuropathy and occurs for several reasons. Symptoms of a peroneal neuropathy are often noticed first on wakening from a normal night’s sleep, probably the result of sleeping in an abnormal position causing nerve compression. During long aeroplane, train and car journeys, the traveller may sleep or sit in such a position that the nerve becomes compressed. Bedridden patients often develop this neuropathy, probably due to a combination of weight loss and pressure on hard hospital mattresses or bed railings. Comatose patients can also lie in such a way as to compress the nerve. Plaster casts unfortunately remain a common cause of peroneal neuropathy. These include below knee casts with a hard upper edge that compresses the nerve as it crosses the fibular neck, and also above knee casts. Leg braces with the upper edge just below the knee, and tight bandages around the knee are other causes of compression at the fibular neck. • Crossing the legs has long been alleged to cause peroneal neuropathy by compressing the nerve between the head of the fibula and the patella or lateral femoral condyle of the opposite leg (fig 5). Leg crossing is such an ubiquitous habit that its role in causing peroneal palsy is difficult to assess, but I believe it to be the most common cause of otherwise unexplained peroneal neuropathy. This is based on the fact that many of these patients admit to being habitual leg crossers; the nerve involved is in the leg they habitually cross over the top of the other; and they invariably recover when they stop leg crossing. Sometimes the patient recounts an episode of unusual pressure; for example, an episode of sitting (or falling asleep) for a long period with the legs crossed and wedged under a table top, or of another person sitting on the patient’s knee when the legs were crossed. Some people, when sitting cross-legged, may either ignore the resulting paraesthesias or are unaware of them because of alcohol, drugs, illness or sleep. Perhaps some do not have warning paraesthesias, or these are not enough to wake them. Recent weight loss is sometimes very clearly

-26-

associated with the development of a peroneal neuropathy (“slimmer’s palsy”). Although a metabolic cause for this condition has been propounded, it is caused by the reduction in the protective padding over the nerve and the satisfaction of once again being able to cross the legs (which obese persons cannot do). • Farm labourers and other workers such as carpet layers who squat or kneel for long periods are particularly at risk of developing peroneal palsies (“strawberry pickers’ palsy”) (figs 6 and 7). When the fully flexed knee is bearing the whole weight of the body, the peroneal nerve is probably compressed between the biceps tendon above and the lateral head of the gastrocnemius and the head of the fibula below or, possibly, this position kinks and compresses the nerve within the fibular tunnel (fig 3). Compression at this site has been found on surgical exploration in affected farm workers. In other related postures there may be additional direct pressure against the nerve when the knee is both flexed and pressed against the ground. • Perioperative peroneal neuropathy not due to direct surgical injury (see above) has been recognised for decades, but is infrequent compared to ulnar neuropathies and brachial plexopathies. The incidence of peroneal neuropathy related to operations remote from the leg can be judged from the report of 421 patients undergoing cardiac bypass surgery: 8 (2%) developed peroneal neuropathies. Compression from leg positioning or leg supports may be the cause, but nerve compression pre- and postoperatively can also occur, as with perioperative ulnar neuropathies, because bedridden patients may lie in such a way to compress peripheral nerves. Some perioperative peroneal neuropathies are associated with the lithotomy position. In a prospective study of 991 adults undergoing general anaesthesia and surgery while positioned in lithotomy, 15 (1.5%) developed lower limb neuropathies. The peroneal nerve was involved in 3 patients (0.3%). The symptoms were sensory only, and in 2 patients the neuropathy was bilateral; they all recovered well. • Postpartum foot drop is usually due to a common peroneal neuropathy, but other causes include an L5 radiculopathy and damage to the lumbosacral trunk (see below). Pressure on the nerve at the neck of the fibula by knee supports is one likely cause. In countries where prolonged natural childbirth is common, such neuropathies are often bilateral and may be due to prolonged squatting. Other reports confirm squatting as a cause, but also point to pressure from the patient’s hands against the nerves while prolonged hip and knee flexion is maintained by the patient herself during labour. • Masses: the most common of these is a ganglion arising from the superior tibiofibular joint. Although benign, they are infiltrative and can compress or invade the nerve. Baker’s cysts may also compress the common peroneal nerve, and sometimes also the tibial nerve. Schwannomas and neurofibromas can arise anywhere along the course of the common peroneal nerve, or its two major branches, but are most common in the popliteal fossa. Other rare mass lesions are listed in the table. Callus

-27-

from old fibular fractures, osteomas, and malignant tumours arising from the head or neck of the fibula are all possibilities. • Several cases of true entrapment of the common peroneal nerve in the fibular tunnel have been confirmed surgically. The characteristic operative findings have been described as a “tight crescentic band at the origin of the peroneus longus ... constricting the nerve which was swollen proximally”. Division of these bands is effective in relieving the symptoms. Such true peroneal entrapment is rare: Sidey explored 26 common peroneal nerves in 23 patients, and of the 8 with no apparent cause for the neuropathy, he found evidence of entrapment in only one. One other report of allegedly surgically proven peroneal nerve entrapments lacks credibility. Cadaver dissection studies have shown that few people have a firm fibrous arch as part of the fibular tunnel; this may explain the rarity of nerve entrapment at this site. Spontaneous entrapment of the peroneal nerve should be considered when there is absolutely no identifiable cause, when other more proximal lesions have been excluded, and when symptoms and signs worsen progressively, and imaging studies (see below) are normal.

• Mononeuropathy multiplex syndromes may involve the common peroneal nerve. These disorders include diabetes, vasculitis, and hereditary neuropathy with liability to pressure palsies. In tuberculoid and borderline leprosy, the peroneal nerve is one of the most frequently involved major peripheral nerves.

• Idiopathic peroneal neuropathy is the frustrating appellation for those patients in whom no specific cause can be found. I believe that with the increasing recognition of leg crossing (including recent weight loss) as a cause, and that with improved imaging techniques, this group is shrinking. Some of these patients may have had an unnoticed episode of external compression, or pressure on the nerve during sleep.

-28-

Figure 2.5This patient sat like this through most of the history taking and denied that he habitually crossed his legs! His foot drop is shown in figure 4. Informed consent was obtained for publication of this figure.

Figure 2.6This young man worked all day tiling a bathroom floor (A), and at the end of the job had a marked left foot drop (B). External compression and kinking of the peroneal nerve was the cause of his peroneal neuropathy. Informed consent was obtained for publication of this figure.

-29-

Figure 2.7This young woman was in the habit of putting on her make-up sitting like this. She would also often read in the same position. After one particularly long period spent putting on make-up she developed asymmetrical peroneal neuropathies. The mechanism is likely to be kinking of the nerves during prolonged knee flexion. Informed consent was obtained for publication of this figure.

Deep peroneal nerve When a patient has neurological deficits restricted to the deep peroneal nerve, the lesion may lie in

that major terminal branch of the common peroneal nerve, or there may be a partial lesion of the common peroneal nerve that only damages the fascicles going to form the deep peroneal nerve.

The anterior (tibial) compartment syndrome results from raised pressure within the fascial compartment that contains the deep peroneal nerve and the muscles it supplies. Causes include excessive exercise, soft tissue trauma, fractures, haemorrhage, occlusion of the anterior tibial artery or its parent trunk, or restoration of blood flow after acute arterial insufficiency in the leg. Severe anterior lower leg pain, swelling and redness are associated with motor and sensory dysfunction of the deep peroneal nerve. The neuropathy is caused by compression by the swollen muscles, and improves rapidly following urgent surgical decompression of the anterior compartment.

-30-

Chronic deep peroneal neuropathy can result from compression by ganglia, osteochondromas and aneurysms. A chronic compartment syndrome has been described, but the involvement of the peroneal nerve is uncertain.

2.2.4 EXAMINATION OF THE PATIENT WITH FOOT DROP A patient with a complete lesion of the common peroneal nerve has the classical clinical picture of

paralysis of dorsiflexion and eversion of the foot, and of extension of the toes, resulting in a foot drop and a characteristic slapping gait. Weakness is present in the foot and toe dorsiflexors, the foot everter muscles, but nowhere else. The important other muscles to test in a patient with foot drop are those supplied by the L4–S1 roots, lumbosacral plexus, and sciatic nerve: the gluteal and hamstring muscles, gastrocnemius, and the tibialis posterior. This last muscle is the foot inverter and is innervated by the same roots as the tibialis anterior (L4, 5), but via the tibial nerve.

The sensory loss of a peroneal neuropathy classically extends over the anterolateral surface of the lower leg and the dorsum of the foot and toes. However, except in total nerve lacerations, the sensory loss is often less widespread than the textbook description and is usually restricted to the dorsum of the foot and some toes, as discussed above (fig 4). Sometimes there are no sensory symptoms or signs. The knee and ankle reflexes should be normal in a peroneal neuropathy. If the ankle reflex is abnormal, then an L5 radiculopathy with involvement of the adjacent S1 root, or a plexus or sciatic nerve lesion are to be considered.

The course of the common peroneal nerve should be examined carefully. When present, local tenderness of the nerve and a Tinel’s sign at the fibular head and neck are valuable indicators of a nerve lesion there; however, one sometimes sees patients with L5 radiculopathies and tenderness of the peroneal nerve. The simple straight leg raising test, when abnormal, is always useful in indicating a root or other proximal nerve lesion. Cysts or other masses in the popliteal fossa or lateral to the knee may be found on careful palpation. Thickening of the nerve occurs in leprosy.

Patients with foot drop sometimes pose a difficult diagnostic challenge for two reasons: first, partial involvement of the common peroneal nerve may produce highly variable degrees of weakness and sensory loss in the muscles and skin supplied by the nerve, and second the nerve lesion lies more proximally than the common peroneal nerve.

-31-

Partial peroneal neuropathies Variable involvement of the muscles and skin supplied by the common peroneal nerve is frequent

in all but complete lacerations of the nerve. Patients often have sensory only or motor only symptoms and signs. The former may involve the superficial or deep sensory branches only. Likewise, the weakness may be confined mainly to the muscles supplied by the deep or the superficial peroneal nerves. This variable motor and sensory involvement is explained by the microscopic anatomy of the nerve. At the knee, the fibres that form the deep and superficial peroneal nerves clearly lie in separate fascicles. Damage to the common peroneal nerve from whatever cause can produce differing involvement of individual fascicles in the nerve similar to the partial deficits seen in radial “Saturday night” neuropathies, and ulnar neuropathies at the elbow.

2.2.5 DIFFERENTIAL DIAGNOSIS OF FOOT DROP The more proximal focal neuropathies that produce foot drop and masquerade as common peroneal

neuropathies are:

• L5 radiculopathies

• Lumbar plexopathies

• Lesions of the lumbosacral trunk of the lumbosacral plexus

• Sciatic neuropathies

The cause may be clear from the history—for example, a patient with characteristic symptoms of lumbar radiculopathy, a hip fracture, a difficult labour that may have involved the use of forceps or a clear episode of compression or trauma to the sciatic or common peroneal nerve. However, in other situations, it may not be so obvious. Lumbar radiculopathy can occur without low back pain that radiates down the leg. Foot drop following hip arthroplasty is much more likely to be due to sciatic nerve damage than to pressure on the common peroneal nerve at the fibular head. Foot drop following delivery of a child may be due to a disc herniation and L5 radiculopathy, compression of the lumbosacral trunk or pressure on the peroneal nerve by knee supports.

An L5 radiculopathy, lumbar plexopathy and a lumbosacral trunk lesion will all produce weakness in muscles not supplied by the common peroneal nerve, notably the gluteal, hamstring and, particularly, the tibialis posterior muscle. However, damage to the lateral trunk of the sciatic nerve can exactly mimic a

-32-

peroneal neuropathy because this trunk becomes the peroneal nerve. Although a trunk lesion involves the innervation to the short head of the biceps femoris, this muscle cannot be clinically examined separately from the other hamstring muscles, so electromyographic examination is required to sort this out. Fortunately, in many cases of lateral trunk lesions, there are often signs of mild involvement of the medial trunk, and these can be revealed by careful clinical and electrophysiological examination.

Sometimes a generalised peripheral neuropathy appears superficially like bilateral peroneal neuropathies because foot drop is more obvious than plantar flexor weakness. In amyotrophic lateral sclerosis, an early feature may be foot drop. And patients with myopathies, particularly the rare distal ones, will often have foot drop. Patients with upper motor neuron weakness of the leg generally have more weakness of foot dorsiflexion than plantar flexion. In all of these, a careful examination will usually show that there are signs beyond the innervation of the common peroneal nerve. Finally, focal dystonia of the foot may present with apparent weakness in dorsiflexion.

2.2.6 INVESTIGATIONS

Nerve conduction and electromyographic studies These are a valuable extension of the clinical examination in evaluating patients with foot drop.

They should be done whenever there is the slightest concern, following a careful examination, that the foot drop may be due to something other than a peroneal neuropathy. Conduction abnormalities at the level of the fibular neck/head are often found, and electromyographic (EMG) abnormalities are restricted to muscles supplied by that nerve. When more proximal lesions are suspected, then a very useful strategy is to do needle EMG examinations of lumbar paraspinal, gluteal and hamstring muscles. Often the two most important muscles to test are the short head of the biceps femoris and the tibialis posterior, for reasons explained above. If searching for evidence of medial trunk of sciatic nerve damage when a sciatic neuropathy with predominant lateral trunk damage is suspected, sural and tibial nerve conduction studies are helpful.

The electrophysiological findings are also useful in predicting recovery. Patients with conduction blocking and little or no axonal damage as determined by EMG study of the tibialis anterior muscle often recover in a few weeks if further pressure on the nerve is avoided. Severe axonal damage neuropathies recover slowly and partially over many months. Patients with mixed axonal and conduction block neuropathies will often have a biphasic recovery: improvement within weeks is due to recovery of those

-33-

fibres that have been affected by demyelination, while the much slower process of axonal regeneration takes many months—up to a year.

Imaging Imaging of the knee should be done when a patient with a peroneal neuropathy of no apparent cause

is not improving, or is worsening. Plain radiographs sometimes reveal a soft tissue mass or bone lesion. Ultrasonography is effective in outlining Baker’s cysts, aneurysms and ganglia (fig 8). CT scanning is excellent for detecting large soft tissue masses in this area. MRI is even more effective for demonstrating the range of intrinsic and extrinsic mass lesions involving the peroneal nerve (fig 9).

Figure 2.8(A) Longitudinal ultrasonographic view of a normal peroneal nerve (PN) and the adjacent fibula (F). (B) Similar view of a peroneal nerve with an intraneural ganglion. (Reproduced with permission from Visser LH. High-resolution sonography of the common peroneal nerve: detection of intraneural ganglia. Neurology2006;67:1473–5.)

-34-

Figure2. 9Proton density-weighted MR axial image of the leg below the knee. The tibia lies superiorly and to the left, the fibula lower and to the right. Lateral to the neck of the fibula is a large mass (arrow) exactly where the common peroneal nerve lies. The patient had a peroneal neuropathy. The lesion was excised and was a neuroma arising from a branch of the nerve. Reproduced with permission from Loredo R, Hodler J, Pedowitz R, et al. MRI of the common peroneal nerve: normal anatomy and evaluation of masses associated with nerve entrapment. J Comput Assist Tomogr 1998;22:925–31.

Case 1 A 69-year-old retired nurse was out for a vigorous walk when she realised that her right foot was

slapping the ground. There was also tingling on the dorsum of the foot and ankle. Over the next several days she tripped on the toes of that foot. She had been dieting and exercising and had lost 22 kg. She sat during the consultation with her right leg crossed over the left, and when this was commented on, she said she had been doing this “all the time” since she had lost weight and was able to cross her legs again. Examination showed weakness of the right tibialis anterior (2/5), extensor hallucis longus (3/5), and the peroneal muscles (3/5). The tibialis posterior and all other muscles in both legs were normal. There was no sensory loss. The peroneal nerve was tender to palpation at the fibular neck. Motor nerve conduction studies showed conduction blocking and slowing at the neck of the fibula; the peroneal sensory response was very small

-35-

compared to that on the left. The patient refused an ankle foot orthosis. She was counselled not to cross her knees. In two months, her strength was somewhat better, and by four months, it was back to normal

Take home messages • Peroneal neuropathy can result from habitual leg crossing and, in this patient, illustrates the phenomenon of

“slimmer’s palsy”.

• The motor deficits are perfectly in keeping with the distribution of the peroneal nerve, but there was no sensory loss presumably because the sensory fascicles within the nerve were not sufficiently damaged. Thetingling was probably due to nerve irritation and abnormal impulse generation.

• Focal nerve tenderness at the knee is a simple but valuable localising sign.

• Merely avoiding leg crossing can be curative.

Case 2 A 33-year-old, very athletic man competed in a long distance run and cycle ride. Two days later, he

realised he had started to limp and when he ran, he tripped over his left foot. He also had a foot drop and numbness over the anterolateral aspect of his lower left leg and dorsum of the foot. He denied any low back or leg pain, and habitual leg crossing. On examination two weeks later, the left tibialis anterior and extensor hallucis muscles were weak (4−/5); however, tibialis posterior, gastrocnemius, hamstring, and gluteal muscles were also weak at 4–4+/5. There was sensory loss over the anterolateral aspect of the lower leg and dorsum of the foot. The left ankle reflex was less brisk than the right. Straight leg raising was normal, and the peroneal nerve was not tender at the fibular neck. Nerve conduction studies showed no evidence of a peroneal neuropathy. Needle electromyographic studies showed fibrillation potentials and polyphasic motor unit potentials in the left tibialis anterior, tibialis posterior, and gluteus medius muscles. A CT scan of the lumbosacral spine showed a large L5/S1 disc herniation on the left compressing the L5 and S1 nerve roots. The patient underwent a discectomy and made a complete recovery.

Take home messages • The patient had symptoms that strongly suggested peroneal neuropathy, but the examination clearly showed

deficits best explained by L5 and S1 radiculopathies.

• In spite of having a large disc herniation, he had no low back pain or “sciatica”.

-36-

• Because of the athletic event preceding the symptoms, an anterior compartment syndrome was an important consideration but the absence of shin pain, swelling and hyperaemia argued strongly against that diagnosis.

Case 3 A 35-year-old primigravida had marked low back pain in the last three months of her pregnancy.

She was admitted at 41 weeks gestation and labour was induced under epidural anaesthesia. The fetus was in the right occipital lie. The second stage of labour was prolonged. Forceps were used to rotate the fetus to an occiput anterior position and then to assist in delivery. Throughout labour the mother experienced considerable pain down the left leg. This persisted and developed into burning paraesthesia over the anterolateral lower leg and dorsum of the foot. She had a marked left foot drop. Examination three weeks later showed marked weakness of the tibialis anterior, extensor hallucis longus, the peroneal muscles, and tibialis posterior muscle. There was mild weakness of the left gluteal muscles and hamstrings. There was decreased light touch over the dorsum of the foot. Reflexes were normal. Electrophysiological studies showed no peroneal motor conduction abnormalities at the knee. The peroneal and sural sensory amplitudes were both reduced in size. Needle EMG showed fibrillations and neurogenic motor unit potentials in the clinically weak muscles, but not in lumbar paraspinal muscles. A CT scan of the lumbar spine was normal. A diagnosis of damage to the lumbosacral trunk of the left lumbosacral plexus was made. She was fitted with an ankle foot orthosis, and made a slow but nearly complete recovery.

Take home messages • here includes L5 radiculopathy from an L4/5 disc herniation, root damage from The differential diagnosis

misplaced epidural injections, and peroneal neuropathy at the knee due to compression against stirrups, or prolonged knee flexion during delivery. Less frequent than all of these is crush damage to the lumbosacral trunk against the ala of the sacrum (fig 1) from the fetal head and/or forceps.

• Careful clinical examination and electrodiagnostic studies can usually localise the lesion.

Imaging of the lumbar spine is helpful when disc herniation is suspected. •

Case 4 A 74-year-old man underwent a right total hip replacement for advanced osteoarthritis. When

mobilised by the physiotherapist two days later, it was apparent that he had a right foot drop, with paraesthesia and numbness over the anterolateral aspect of the lower leg and dorsum of the foot. He went on to develop allodynia in this area and neuropathic pain in the lower leg. The anaesthetist was blamed for

-37-

poorly positioning the leg and causing a peroneal neuropathy and the orthopaedic surgeon was contemplating exploring the peroneal nerve at the knee. However, although careful neurological examination confirmed major weakness of the peroneal innervated muscles, and the gluteal, hamstring and tibialis posterior muscles were not weak, the gastrocnemius was mildly weak and the ankle reflex was absent. Also, the sensory abnormality was not only mainly in the peroneal nerve distribution, but there was some impairment in the lateral border and sole of the foot. Nerve conduction studies showed no evidence for a peroneal motor conduction abnormality at the knee. Both the peroneal and sural sensory potentials were abnormally small. Needle EMG studies showed fibrillations and neurogenic abnormalities in the tibialis anterior, the short head of the biceps femoris and, to a lesser degree, in the gastrocnemius. A diagnosis of a lesion of the sciatic nerve in which the predominant damage was to the lateral trunk, was made. The patient was fitted with an ankle foot orthosis, given medications for neuropathic pain, and eventually regained considerable strength.

Take home messages • e involving exclusively or predominantly the lateral trunk closely mimics a

eferentially or exclusively damage the lateral trunk.

eals to that trunk as well.

branch of the lateral trunk, is particularly useful.

Most patients with a peroneal neuropathy fall into one of three main groups:

•

• ssive episode, or recurrent episodes

•

-38-

A lesion of the sciatic nervperoneal neuropathy.

• Many traumatic lesions of the proximal sciatic nerve pr

• Careful examination of the muscles and skin supplied by the medial trunk of the sciatic nerve often revsome damage

• Electrodiagnostic studies are a valuable tool in diagnosing the site of the lesion. Needle EMG of the short head of the biceps femoris muscle, innervated by the only Nerve conduction studies of the tibial nerve and EMG of muscles innervated by it are helpful in confirming some damage to the medial trunk of the sciatic nerve.

2.2.7 MANAGEMENT OF THE PATIENT WITH FOOT DROP

acute trauma

a recognisable compre

progressive neuropathy of uncertain cause.

If the type of trauma—for example, a laceration, suggests a complete transection, immediate nerve repair is indicated. Incomplete traumatic neuropathies are followed clinically, and spontaneous improvement

h progressive peroneal neuropathies should have imaging studies. The choice between ultrason

ass

,

-39-

is expected. If this does not occur within several mo surgical exploration of the nerve may be warranted because scar tissue and neuroma formation can impe xonal regeneration; careful surgical resection of these, combined with cable nerve grafting can often bring about some degree of recovery. For blunt injuries that appear to be complete, waiting for spontaneous improvement to occur over a period of several months is probably the best approach. Stretch injuries of the peroneal nerve are often associated with long areas of damage. Even when they are repaired with nerve grafts, the results are poor. The acute compartment syndromes of the lower leg are surgical emergencies. Prompt fasciotomy is important for good recovery of nerves and muscles.

When a definite compressive episode can be identified I re-evaluate the patient in two months. When leg crossing or other postural habits that may repetitively compress the nerve are suspected, I instruct

nths,de a

the patient to avoid these and re-evaluate in two months. Extensive investigations are usually not necessary. If a patient does not show signs of improving at this time, or if they worsen, they are categorised in the progressive group.

Patients witography, CT scanning and MR imaging depends on local availability, expertise, and cost

considerations. I usually do ultrasound first, then CT if the diagnosis is still uncertain, and then MRI if necessary. If a mass is discovered, it should be surgically removed unless it is a bony metastasis. If no mis found, surgical exploration is still warranted. Some soft tissue lesions may not be visible on imaging (withimproved imaging this is becoming less likely). If the patient has a true entrapment of the nerve within the fibular tunnel, this can only be confirmed surgically.

Patients in whom a popliteal mass is palpated when they first present with foot drop should be sentfor imaging. For patients with suspected radiculopathy, imaging with CT or MR should be undertaken and further management then decided. If a lumbosacral plexus lesion is suspected, then CT or MR imaging of the lower lumbar retroperitoneal area and the pelvis should be done. For sciatic nerve lesions with foot dropit may not be necessary to perform imaging studies—for example, the post-hip surgery patient. However, if there is no apparent cause, then the main trunk of the sciatic nerve from the sciatic notch to the popliteal fossa should be imaged with MR using gadolinium enhancement searching for tumours of the nerve, or other lesions that may be impinging on it.

Regardless of the cause of the peroneal palsy, patients with foot drop benefit greatly from a brace to support the foot, aid walking and prevent tripping. The most satisfactory is a lightweight plastic ankle foot orthosis that fits inside the shoe and up the back of the calf (fig 10). In cold climates, a pair of boots often suffices. If the foot drop is permanent, surgical treatment with tendon transfers should be considered.

Figure 2.10(A) Foot drop caused by a severe L5 radiculopathy. (B, C) Ankle foot orthosis in place. Informed consent was obtained for publication of this figure.

Practice points • The most common cause of a unilateral foot drop is peroneal nerve injury.

• Most of these patients have had a single episode of pressure to the nerve such as during a long aeroplanejourney, are habitual leg crossers, or have an occupational cause for the neuropathy.

• ve

• When such causes are identified, the patient should be advised to avoid such risks, and good recovery is to be expected.

• Proximal nerve lesions mimicking a peroneal neuropathy include L5 radiculopathy, lumbosacral plexopathy and sciatic neuropathy; these produce weakness in muscles outside of the peroneal territory—for example, the hamstring and the foot invertor (tibialis posterior) muscles, and sensory loss in the tibial nerve territory.

• EMG and nerve conduction studies are helpful in differentiating proximal lesions from peroneal neuropathy.

• Footdrop due to an L5 radiculopathy sometimes occurs without any low back or radicular pain.

Imaging is helpful in diagnosing the cause of foot drop, be it a more proximal nerve lesion or a progressiperoneal neuropathy at the knee.

-40-

• Whatever the cause of the foot drop, it is important to prescribe an ankle foot orthosis to improve gait and help prevent falls.

2.3 Functional electrical stimulation of walking: Function, exercise and rehabilitation

2.3.1 Abstract For nearly half a century, functional electrical stimulation (FES) has been used to restore walking

for people with paralysis and muscle weakness due to stroke and spinal cord injury. The first applications of the technology were intended to permanently replace lost neuromuscular function. Later, FES-assisted walking was found to have therapeutic benefits that include increased muscle strength, cardiovascular fitness and improved gait function that could be maintained after use of FES was terminated. In this review, we examine some of the major FES-assisted walking systems that have been developed for experimental and commercial purposes over the last four and a half decades, including foot drop stimulators, multichannel stimulators and hybrid orthotic systems. Keywords: Functional electrical stimulation (FES); Gait; Walking; Spinal cord injury; Stroke

2.3.2 Introduction Functional electrical stimulation (FES) is the application of electrical impulses to neuromuscular

tissue to restore useful movements. Since the 1960s, FES has been used to evoke involuntary muscle contractions and help people from disabled populations stand and walk. Today, this technology is used for both functional and therapeutic purposes. A variety of simple FES-assisted walking devices have been developed to help people with hemiplegia due to stroke to overcome muscle weakness in their affected leg. In spinal cord injury (SCI), various forms of FES-assisted gait have been shown to enhance gait, muscle strength and cardiorespiratory fitness. Several dozen FES devices have been developed and tested over the last four and a half decades, and many have been commercialized and made available for everyday use at home and in the community. Some applications are specifically intended for therapeutic use in clinical settings while others are used as permanent orthotic systems.

-41-

2.3.3 Brief history of FES-assisted walking The first applications of FES were designed to restore lower limb function in patients who had

experienced a stroke or a SCI. In 1960, Kantrowitz demonstrated paraplegic standing by applying continuous surface FES to the quadriceps and gluteus maximus muscles of a patient with complete SCI. Stimulation of thesemuscle groups produced full extension of the lower limbs, and enabled a form of gait similar to walking with long leg braces. Around the same time, Liberson et al. developed a simple FES device to correct foot drop, a common symptom in hemiplegia characterized by a lack of dorsiflexion during the swing phase of gait which results in short, shuffling strides. Liberson's device, which has the distinction of being the first FES device to receive a patent, consisted of a power supply worn on a belt, twoskin electrodes positioned for stimulation of the common peroneal nerve, and a heel switch. The stimulation was activated whenever the heel lost contact with the ground, and it was deactivated when the heel regained contact with the ground. The majority of FES-assisted walking devices in use today employ one or both of the techniques introduced by Kantrowitz and Liberson almost 50 years ago.

Several technological advancements were made in the decades following Kantrowitz and Liberson that allowed for more sophisticated control of walking. As microprocessor technology developed, FES devices became more portable and flexible. Microprocessors also led to the development of multichannel stimulators, which are FES devices that deliver different signals to more than a single stimulation site. Multichannel stimulators are capable of controlling complex, coordinated movements of the lower limbs. One such system was a six-channel stimulator developed at the University of Ljubljana in Slovenia. This device synchronized stimulation to up to six different muscle groups to produce gait patterns in patients with complete paraplegia.

Another important technological advance was the development of percutaneous and subcutaneous electrodes. Percutaneous electrodes are long needle-like electrodes that are inserted through the skin and into the muscle tissue. They are implanted temporarily and removed when the specific activity is complete. Compared to surface stimulation systems, percutaneous electrodes provide more precise control of specific muscles, provide access to deeper muscles and consume less electrical energy. Subcutaneous electrodes are surgically implanted into muscle or nerve tissue. These electrodes are usually part of a fully-implanted system that includes a power source, a microprocessor and a coil for communicating wirelessly with components outside of the body.

-42-

2.3.4 FES-assisted paraplegic gait In the 1980s, Kralj et al. described the basic technique for paraplegic gait using surface stimulation.

According to Kralj's technique, four channels of stimulation are used. Skin electrodes are placed over the quadriceps muscles and peroneal nerves bilaterally. The user controls the FES device with two pushbuttons attached to the left and right handles of a walking frame, canes or crutches. When the FES device is activated, both quadriceps muscles are stimulated. The left button initiates swing phase in the left leg by briefly stopping stimulation of the left quadriceps and stimulating the left peroneal nerve. This stimulation is applied suddenly so as to trigger the flexor withdrawal reflex, resulting in simultaneous hip and knee flexion as well as dorsiflexion. After a fixed period of time, peroneal nerve stimulation is stopped and quadriceps stimulation is resumed. Similarly, the right button initiates swing phase in the right leg. Kralj et al. successfully applied this system to several dozen subjects with SCI.

Kralj's basic technique is still one of the most popular in use today. It is used, for example, by the Parastep-I® system. This is the only FES device for walking to receive approval from the United States’ Food and Drug Administration (FDA). It is an elaborate surface FES device that is intended to be used by people with complete or incomplete paraplegia every time they wish to walk. It includes a two-wheeled walker and an ankle-foot orthosis to bolster ankle stiffness. The Parastep-I® is commercially available, and over 600 people have used it successfully. A major disadvantage of systems that use the Kralj technique is that they rely on the flexor withdraw reflex, which is variable and subject to habituation.

More complex multichannel systems have been developed to enable better muscle specificity and control in paraplegic gait. There have also been many advances in the development of automatic control of FES-assisted gait, however, manually triggered open-loop control still remains practical and most popular with users. In all its forms, FES-assisted paraplegic gait tends to be slow, awkward, unnatural looking and requires a great deal of energy from the user. The high energy cost is a major limitation for neuroprostheses that are intended to be used as permanent assistive devices. However, when FES-assisted walking is used to facilitate exercise, inefficiency is actually a desirable feature since the user has to provide substantial effort to be able to walk with the system.

One major limitation of FES-assisted walking devices based on surface stimulation is that the hip flexors cannot be stimulated directly. Therefore, hip flexion during walking must come from voluntary effort, which is often absent in paraplegia or the flexor withdrawal reflex (initiated by peroneal nerve stimulation).

-43-

2.3.5 Fully implanted systems Fully implanted systems offer several advantages over surface and percutaneous systems for long-

term usage, including improved convenience, cosmesis and the ability to stimulate deep muscles such as the hip flexors. As a result, they can provide greater muscle selectivity and more natural gait patterns. Two such systems are the Praxis FES system and the system proposed by Kobetic, which use 24 and 32 electrodes, respectively. The Praxis24 system also provides bladdervoiding. It is important to note that these systems have only been implanted in a small number of complete paraplegics. Reportedly, they enable a swing-through gait pattern, much like long-leg brace walking, which is not a reciprocal walking pattern.

2.3.6 Foot drop stimulators Foot drop stimulators (FDS) are a special class of FES devices that specifically address the problem

of foot drop, which is a lack of dorsiflexion due to weakness of the tibialis anterior muscle. Most FDS are simple, single-channel systems that are intended for everyday use, such as Liberson's invention in 1961. They are used mostly by people with hemiplegia due to stroke, but are also used bilaterally by some people with paraplegia.

Most FDS apply stimulation to the common peroneal nerve, which causes contraction of the muscles responsible for dorsiflexion (i.e. tibialis anterior, extensor hallucis longus, etc.). It can also trigger the flexor withdrawal reflex, which may not be desirable. The flexor withdrawal reflex occurs naturally when a sudden, painful sensation is applied to the sole of the foot. It results in flexion of the hip, knee and ankle of the affected leg and extension of the contralateral leg in order to get the foot away from the painful stimulus as quickly as possible.

Some FDS have been commercialized, for example the MikroFES (Josef-Stefan Institute, Ljubljana, Slovenia) and the Odstock Dropped Foot Stimulator. The latter was shown to significantly increase walking speed and efficiency when used, and a carry-over effect was observed in stroke patients; that is, their walking speed and efficiency without the stimulator was improved. Similar studies have reported no carry-over effect. Users of the Odstock device were generally satisfied with it, however almost all of them identified the surface electrodes as problematic, and two thirds would consider an implanted system instead. The BIONic Walkaide is a new generation FDS that uses an implantable microstimulator, which provides more selective control of muscle activation.

The first commercially available implanted FDS was developed by Rancho Los Amigos Medical Centre and Medtronic, Inc. . The surgically implanted components were a radio-frequency (RF) receiver,

-44-

pulse train generator, and one bipolar electrode implanted adjacent to the peroneal nerve. An external unit worn on the belt delivered power via the RF coil and received input commands from a wireless foot switch. Despite some problems with electrode migration and infection, the device was considered successful. Since then, more reliable and easier to implant systems have been devised, such as the Neurostep system and the ActiGait implanted FDS, but they are not yet commercially available. The latter uses input from an implanted cuff electrode around the sural nerve, which is the nerve innervating the skinsensors on the sole of the foot. This system is unique in that it requires no external sensors.

Gait detection is extremely important for the success of a FDS and any FES-assisted walking device. Stimulation sequences must be properly synchronized with the kinematic events of gait, such as swing phase and stance phase. Most modern FDS use a heel switch on the affected leg for detection of stance phase. Burridge et al. tried using the foot switch on the non-affected leg, but found it was not preferable unless the patient was unable to reliably achieve heel contact on the affected leg. Vodovnik was one of the first to experiment with manual pushbuttons and EMG sensors. Other alternatives to the heel switch include a heel/toe switch, an array of four single-axis accelerometers positioned on the shank, a tilt sensor positioned on the shank, electroneurography, and an array of force-sensitive resistors and a gyroscope located in the insoles. There is much debate concerning the reliability of these sensors. Robust controller design is necessary to interpret imperfect signals from gait phase detection sensors. To date the sensor proposed by Pappas et al. has shown the highest detection accuracy and reliability in real-life outdoor testing conditions.

2.3.7 Hybrid orthotic systems Devices that combine FES and mechanical braces are called hybrid orthosis systems. One example

is the Parastep-I system, which includes a pair of ankle-footorthoses (AFO) to prevent foot drop. From a biomechanical perspective, the purpose of braces is to improve stability by reducing the number of degrees-of-freedom of the body. Normal walking is a complex motion involving 22 degrees-of-freedom in the pelvis and lower extremities. Hybrid orthoses aim to reduce that number, thus compensating for specific muscle weakness, while applying FES to the muscles that act on the remaining degrees-of-freedom. Hybrid orthotic systems have been created by combining FES with existing bracing systems, such as the Reciprocating Gait Orthosis (RGO) and the Medially Linked Knee-Ankle-Foot Orthosis (MLKAFO). Both of these braces reduce the number of degrees-of-freedom of the pelvis and lower extremities to seven. They address the issue of pelvis instability that is experienced by many individuals with SCI.

-45-

Hybrid orthoses can be an effective clinical option for individuals with significant orthopedic complications that might otherwise contraindicate the prescription of conventional braces or FES alone. Some hybrid orthoses have been shown to increase gait speed, stride length and cadence when compared to FES alone. Hybrid orthotic systems are in use in a small number of laboratories and clinics around the world. However, they have yet to achieve widespread acceptance from clinicians and consumers.

2.3.8 Therapeutic applications Numerous reports over the past three decades have asserted positive therapeutic effects of FES-

assisted walking in incomplete SCI. In this paradigm, FES is applied to the lower limbs, usually via surface electrodes, at sub-maximal levels. While walking, FES-induced contractions are superimposed over the patient's voluntary contractions resulting in greater muscle force, joint range of motion and load bearing in the legs. Typical regimens of FES-assisted walking therapy involve three to five one-hour sessions per week for at least four weeks. Depending on the user, these sessions may take place in the home or in a supervised clinical setting. Often, walking is performed on a treadmill and with a safety harness.

In its many forms, FES-assisted walking has been shown to reduce spasticity, increase bone mineral density, increase voluntary muscle strength and coordination. Furthermore, there is growing evidence that regular use of FES by people with various neurological disabilities can result in recovery of voluntary muscle control and improved function after the stimulator is taken away and. A carry-over effect in terms of overground walking speed has been demonstrated in a large population of incomplete SCI patients using a variety of neuroprostheses for walking.

There is growing evidence that task-oriented locomotion therapy can help the damaged central nervous system adapt and form new neural pathways for voluntary function. In humans, there is evidence that the injured spinal cord continues to reveal a great deal of plasticity even several years after SCI. External electrical stimulation has been shown to affect neural plasticity. Furthermore, a hypothetical mechanism has been proposed in which motor neurons within the spinal cord can modify synaptic organization to restore conduction after an incomplete lesion. When electrical stimulation of motor axons coincides with voluntary muscle effort, restorative modifications in the central nervous system may be induced. This mechanism may account for the improvements in walking function observed in several recent studies on the efficacy of FES-assisted walking as a clinical intervention for incomplete SCI and FES-assisted walking has also been shown to improve function in children with spastic cerebral palsy.

-46-

In hemiplegia due to stroke, FES-assisted walking has had some success, but evidence of its efficacy is very limited.

2.3.9 Limitations FES-assisted walking has experienced many technological advances over the decades, including

multichannel stimulation, implantable electrodes, advanced sensors and control systems. However, it is still widely perceived as an experimental procedure that has not been proven effective in a large portion of the general patient population. Its effectiveness is very dependent on the individual's specific condition. For example, patients who have denervation muscles due to peripheral nerve injuriesare not candidates for FES.

One of the major limitations of FES is that it tends to produce rapid fatigue in muscles. This is due to the asynchronous recruitment of muscle fibers, and it greatly limits the duration for which FES-assisted walking can be performed by many patients. In many cases, when voluntary muscle force is low and high intensity FES must be used, muscle fatigue is apparent after 1 min of stimulation. The muscles can be too fatigued to produce gait movements a few minutes later. Randomization of stimulation parameters has been recommended as one method to reduce the fatiguing effect of FES, however, its efficacy lacks evidence.

2.3.10 Conclusions FES is one of the most effective ways to improve gait in individuals after stroke or SCI. Both

implanted and surface systems have demonstrated the utility and efficacy of FES time and again. Today, the most promising method of applying FES technology to improve gait is FES-assisted walking therapy and the best candidates for this therapy are stroke and incomplete SCI patients. Different modalities of the FES therapy have been proposed, and many have been reported efficacious in the literature. Positive results have also been reported for implanted and surface FES devices that are used as permanent orthotic systems. Although these systems have been found efficacious in a number of studies, a limited number of consumers are using these devices at home and in clinical settings. We believe that further technological improvements, patient education, intensive and focused therapy aimed at teaching consumers how best to use the device will play a key role in allowing this technology to reach its full potential. In summary, FES-assisted walking remains one of the most promising rehabilitation techniques to restore gait in stroke and SCI patients. We are confident that in the future FES-assisted walking will be an integral part of stroke and SCI rehab.

-47-

2.4 Matlab/STM32 Program

From RapidSTM32

This example demonstrates how to create a Simulink model, for using FiO Std Evaluation Board (EVB) as a target, to acquire real-time data from two sources, including:

1. A potentiometer 2. An on-chip temperature sensor

Both analog inputs are sampled, and converted to digital values, by an on-chip 12-bits Analog to Digital Converter at 100Hz sampling rate. The converted data are sent to the host PC via USB Human Interface Device (HID) protocol. The host PC receives the data, converts the measured temperature from volts to Celsius, and finally, displays both data, in Simulink as both numerical values, as well as time series plots in real-time.

2.4.1 Watch the Step-by-Step Video

http://www.youtube.com/watch?v=-kPhiYWq8Tg&feature=player_embedded

http://www.youtube.com/watch?v=cj6UI68l5z0&feature=player_embedded

2.4.2 Step-by-Step Guide

Creating a system model for the FiO Target

1. Set FiO Std in USB In Application Programming (IAP) Mode and connect it to the host PC. This is required because, we will be using the Target USB Send Block and the Compile and Download Control Block in our example model, which requires that a FiO board is connected to the host PC to enable the full-feature mode of the RapidSTM32 Blockset.

-48-

2. Test connection between the target and the host PC with request_productinfo('usb','name') command as shown here.

3. Create a new model and set it to use RapidSTM32 target and insert Setup System Clocks & SysTick Block, Compile and Download Control Block, ADC Configuration Block, and Target HID Send Block as shown below. Note that:

• Setup System Clocks & SysTick Block and Compile and Download Control Block are from Device Configuration Library

• ADC Configuration Block and Target HID Send Block are from On-Chip Peripherals Library

Figure 2.11 Overall system model

4. Configure the ADC Configuration Block as shown below. Note that:

• Use Custom Setting

-49-

• Channel Selection [16 13] (Analog CH16 and CH13 are connected to the on-chip temperature sensor and Pin C13 respectively.)

• Output Data Type Volts (single - 32bits) • Sample time (in seconds) 0.01

Figure 2.12 Configure the ADC Block

5. Configure the Target HID Send Block as shown below. Note that:

-50-

• Use 'single single' for Binary data • Enable Blocking • Sample time (in seconds) -1 to inherit the sample time of 0.01 seconds from the previous block.

Figure 2.13 Configure Target HID Send Block

6. Connect all blocks as shown below.

-51-

Figure 2.14 Connected model

7. Save your model as adc_hid_target_demo, so we know clearly that this model is for the target (to run on a FiO board). Then, update the diagram twice. All blocks should display correct sample time of 10 milli second (or 0.01 second).

-52-

Figure 2.15 Updated Model

8. If you have not done so, set FiO Std in USB IAP Mode. Test connection between the target and the host PC with request_productinfo('usb','name') command as shown here.

9. Initiate the Build Process. The source code should be automatically generated, compiled, and downloaded to your target at the end of the process as shown below.

-53-

Figure 2.16 Build target complete

Creating a system model for the Host PC

1. Create a new model and set it to use RapidSTM32 target. Insert Host HID Receive Block, Convert Onboard AN16 Volts to Temp C, Display Block, Scope Block, and Data Type Conversion Block as shown below. Save this model as adc_hid_hostpc_demo.mdl. Note that:

• Host HID Receive Block is from Add-On Modules Library • Convert Onboard AN16 Volts to Temp C is from On-Chip Peripherals Library • Display Block, Scope Block are from Simulink Sinks Library • Data Type Conversion Block is from Signal Attributes Library

-54-

Figure 2.17 Host PC system model

2. Configure the Host HID Receive Block as shown below. Note that:

• Binary data format is 'single single' • Sample time is 0.01 second.

-55-

Figure 2.18 Configure Host HID Receive Block

3. Configure the Data Type Conversion Block as shown below. This is required since the Display Block requires double data type input but the Host HID Receive Block outputs single data type.

-56-

Figure 2.19 Configure Data Type Conversion Block

4. Connect all blocks as shown below.

-57-

Figure 2.20 Completed host PC model

5. Change the final simulation time to inf (infinity) so our host PC system model will run continuously until we press the stop button. Update the diagram twice or until all blocks display correct sample time. We are now ready to proceed with the simulation.

-58-

Figure 2.21 Updated diagram before simulation

Running the system

1. Connect FiO Std potentiometer output to pin C3 as shown below.

-59-

Figure 2.22 Connect FiO Std potentiometer output to pin C3

2. Set FiO Std in Run Custom User Program Mode.

3. Run the host PC Simulink model. Adjust the potentiometer on the FiO Std board and observe the output signals on the Simulink Display and Scope, and oscilloscope (if any).

-60-

Figure 2.23 ADC HID Demo system test

-61-

2.4.3 Download

1. adc_hid_demo.7z 2. adc_hid_demo video part 1 of 2 3. adc_hid_demo video part 2 of 2

2.4.4 Related Examples

1. Displaying ADC data on a character LCD (1 line) 2. Displaying data on a character LCD (multiple lines) 3. Open Loop Example: FiO ADC to Simulink Scope and Simulink Slide bar to PWM 4. Closed Loop Example: FiO ADC to Simulink algorithm to FiO PWM

 

-62-

บทท 3

แนวทางการพฒนา

3.1 การพฒนาซอรฟแวร

การพฒนาซอรฟแวรสวนใหญจะอยในสวนของใชโปรแกรม Simulink แลวนามาเชอมตอกบบอรด

Fio Std โดยพวกผมเลอกทจะใช Microcontroller (MCU) ทงหมด 2 ตว เนองจากวาในตอนแรกพวกผมได

ลองใช MCU ตวเดยวดแลวปรากฏวาผมไมสามารถแสดงขอมลทางหนาจอ LCD ได เนองจาก LCD ไม

สามารถทางานไดตามอตราการชกตวอยางทพวกผมตงไวได ดงนนพวกผมจงตองเพม MCU เขาไปอก 1 ตว

เพอทจะไดทาใหหนาจอ LCD สามารถทางานไดทน

โปรแกรมทจาเปนตองมในเครองคอมพวเตอรประกอบดวย

1. MatlabTM 2009a (v7.8)

2. SimulinkTM (v7.3)

3. Real-Time WorkshopTM (v7.3)

4. Real-Time Workshop Embedded CoderTM (v7.3)

5. Realview MDKTM for ARM (v4.0)

6. Microsoft .NET Framework (v3.5)

3.1.1 ซอรฟแวร ทเขยนใน MCU ตวท 1

หนาทหลกของ MCU ตวแรกคอรบคาจากผใช โดยผใชสามารถกาหนดคาไดเองถง 5 คา

คอความถ ความกวางสญญาณ ระยะเวลากระตน ระยะเวลาหยดกระตน และแอมพลจดสญญาณ

เนองจากผใชแตละคนมความตองการในการกระตนทแตกตางกน บางคนตองการแรงๆในขณะทอก

คนไมตองการแรงมาก ขนอยกบกลามเนอของแตละบคคลดวย อกหนาทหนงของ MCU ตวแรก

คอการสรางสญญาณออกมา 2 สญญาณทเหลอมกนเพอทจะนาสญญาณมาเขาวงจรฮารตแวรตอไป

นอกจากน MCU ตวแรกนยงเปนตวสงการ LED สสมทใชแสดงวาขณะนมสญญาณกระตนอย

หรอไม

-63-

2 1 6

ภาพท 3.1 แสดงภาพรวมของ MCU 1 บนโปรแกรม Simulink

3 4 7

5

สาหรบรายละเอยดตางๆเปนดงน

หมายเลข 1 : เปนกลองทชอวา Digital Output ทาหนาทรบคาสวตซจากผใช นนคอถา สวตซตดอยจะสง

1 เขามา แตถา สวตซปดอยกจะสง 0 เขามา

หมายเลข 2 : เปนกลองทชอวา Data Type Conversion ทาหนาทแปลงคาทสงเขามาใหกลายเปน Boolean

หมายเลข 3 : เปนกลอง Clock และตว Compile ซงตองใสทกครงทใชกลองเกยวกบระบบสมองกลฝงตว

หมายเลข 4 : เปนกลอง ADC Configuration ทาหนาทรบคาจากผใชแลวนามาผาน ADC แปลงเปนคา

ในชวงทเราเซตไว แลวนาไปใชตอไป

หมายเลข 5 : เปนกลอง Digital Output ทใชแสดงผลทาง LED สสม ถาไฟตดแสดงวากระตนอย

หมายเลข 6 : เปนกลอง Digital Output ทใชสรางสญญาณออก 2 สญญาณ ซงจะนาไปตอกบฮารตแวร

ตอไป

-64-

หมายเลข 7 : เปนกลอง Chart หรอ State flow ซงเปนกลองทใชรวมการทางานของ MCU ตวนทงหมด

โดยเราสามารถคลกเขาไปเขยนระบบการทางานของ MCU ในกลองนได โดยภายในของ

กลองนจะเปนดงรป

ภาพท 3.2 แสดงการทางานของ State Flow ใน MCU 1

จากภาพท 3.2 จะเหนวาพวกผมแบงการทางานออกเปน 4 โหมดดวยกน

-65-

1. Calculating mode

เปนโหมดททาหนาทคานวนคาตางๆทรบมาจากผใชซงคาเหลานจะนาไปใชในโหมดอนตอไป

เราสามารถใชโหมดนไดโดยการปด สวตซทกตว

 L2=0.0275/Freq‐Pulse_Duration*590.909090909090909*10^‐6 + 0.0002; 

LED=0; H1=450*10^‐6; 

L3=0.0191; 

Off=Offtime_or_Lockout*18.181818181818181818181818181818; 

On=Ontime_or_Maximum_Stimulation*18.1818181818181818181818181818; 

 du:H=(Pulse_Duration*295.4545454545454545455)*10^‐6; 

Calculating_model 

2. Normal mode

เปนโหมดทสงใหเครองทาการกระตนตลอดเวลา โดยถาเครองทางานในโหมดนไฟ LED ส

สมกจะตดตลอดดวย เราสามารถเลอกใชโหมดนไดโดยการกด SW1

ภาพท 3.3 แสดงการทางานใน Normal mode

จากภาพจะเหนวาวงจร State Flow นแบงการทางานออกเปนชวงกระตนและไมกระตน โดย

S1 คอสญญาณแรกและ S2 คอสญญาณทสองทตองการสราง และการเชอมโยงในแตละกลองกจะ

มเวลาระบไวชดเจน ซงคาเหลานกมาจากคาทผใชระบในโหมดแรก และการออกจากชวงกระตนจะ

-66-

ถกบงคบใหออกในตอนทสญญาณ S1,S2 มคาเปน 0 เสมอ เพอปองกนอนตรายทจะเกดขนกบ

ผใชงาน

3. Manual switch mode

ตงแตโหมดนเปนตนไปจะเหมาะทจะใชกบผปวยโรคเทาตก โดยในโหมดนจะเปนโหมดท

ใหผใชเลอกกดกระตนเอง โดยผใชสามารถเลอกใชโหมดนไดโดยการเปด SW1,SW2 กจะเปนการ

เขาสโหมดน หลงจากนนถาตองการใหเครองทาการกระตนกใหกดสวตซ Hold เครองกจะทาการ

กระตนตลอดเวลา ถาตองการหยดกใหกด Hold อกครงหนง ทสาคญการกด Hold แตละครงกจะม

การเกบคาพารามเตอร Threshold On, Off ซงจะนาไปใชในโหมดถดไปดวย โดยถากดเปด Hold ก

จะเปนการเกบคา Threshold On และถากดปด Hold กจะเปนการเกบคา Threshold Off สวน

รายละเอยดวาเกบไปทาไมนนจะอธบายในโหมดถดไป

ภาพท 3.4 แสดงการทางานในโหมด Manual switch

จากรปจะเหนวาในโหมดนกจะคลายๆกบโหมดทแลวนนคอมทงชวงกระตนและไมกระตน

ทแตกตางกนคอโหมดทแลวพอหยดกระตนไปไดระยะเวลาหนงกจะกระตนตอเอง แตในโหมดน

ผใชจะตองเปนผกดสวตซ เครองถงจะกระตนได ทสาคญในโหมดนพวกผมไดทาการคงคาความถ

ไวท 50 Hz. และคาความกวางสญญาณไวท 500 ไมครอน และใหผใชปรบแอมพลจดไดอยางเดยว

ซงเพยงพอตอการปรบความแรงแลว

-67-

4. Tilt sensor mode

โหมดนจะเปนการกระตนโดยใชคาจากเซนเซอรเปนตววด ผใชสามารถเลอกใชโหมดนได

โดยเปดสวตซ SW1, SW2, และ SW3 โดยเครองจะกระตนกตอเมอคาทไดจากเซนเซอรมคา

มากกวา Threshold On ทไดเกบคาไวในโหมดทแลว และจะหยดกระตนกตอเมอคาทไดจาก

เซนเซอรมคานอยกวาคา Threshold Off ทไดจากโหมดทแลว

ภาพท 3.5 แสดงการทางานในโหมด Tilt sensor

คา Stim_enable ทเหนในรปเปนคาทไดมาจาก MCU 2 ทจะพดถงตอไป สวนวงจรการ

ทางานหลกกจะคลายๆกบในโหมดทผานๆมา

3.1.2 ซอรฟแวร ทเขยนใน MCU ตวท 2

หนาทหลกของ MCU ตวท 2 กคอเกบคาจาก Accelerometer Sensor เพอนาไปใชใน

โหมด Tilt Sensor และการแสดงคาบนหนาจอ LCD โดยในแตละโหมดหนาจอ LCD กจะแสดง

แตกตางกนออกไป ซงรายละเอยดจะกลาวถงตอไป นอกจากน MCU ตวท 2 นยงมการใช LED

อก 3 ส เพอแสดงใหผใชรวาเครองทางานอยในโหมดใด

-68-

1

2

3

4 6 7

5

ภาพท 3.6 แสดงภาพรวมของ MCU 2 บนโปรแกรม Simulink

สาหรบรายละเอยดตางๆเปนดงน

หมายเลข 1 : เปนกลองทใชตง Clock และ Compile ซงไดกลาวถงตอน MCU 1 ไปแลว

หมายเลข 2 : เปนกลองทชอวา ADC Configuration เอาไวรบขอมลจากผใชรวมถงขอมลทสงมาจาก

เซนเซอรดวย โดยแบงเปนรบคาจากผใช 4 ขา และรบคาจาก เซนเซอร3 ขา

หมายเลข 3 : เปนกลมของกลองทใชแปลงคาทรบมาจาก เซนเซอรโดยจะผานกลองตางๆจนไดคาสดทาย

ออกมาเปนมมเอยงทเราตองการ

หมายเลข 4 : เปนกลองทชอวา Data Type Conversion ทาหนาทแปลงคาทสงเขามาใหกลายเปน Boolean

-69-

หมายเลข 5 : เปนกลองทชอวา Digital Output เปนกลองทเอาไวสงคาไปยง LED 3 ตว เพอแสดงการทางาน

ในโหมดตางๆ

หมายเลข 6 : เปนกลมของกลองทใชแสดงผลทางหนาจอ LCD โดยพวกผมไดสรางกลองตางๆเพอทจะให

จอ LCD แสดงผลตามทตองการ ซงมอยท งหมด 3 แบบดวยกน ในโหมด Calculation

และ Normal mode หนาจอ LCD จะแสดงคาความถ ความกวางสญญาณ ชวงกระตนและ

ชวงหยดกระตนออกมา ขณะทโหมด Manual Switch หนาจอ LCD จะแสดงคาความถ

Threshold On, Off ทางหนาจอ สวนในโหมด Tilt Sensor หนาจอ LCD จะแสดงคามมท

เบยงเบน ชวงกระตนและชวงหยดกระตนออกมา

หมายเลข 7 : เปนกลองทชอวา Chart หรอ State Flow ซงเปนกลองทใชรวมการทางานของ MCU ตวน

ทงหมด โดยเราสามารถคลกเขาไปเขยนระบบการทางานของ MCU ในกลองนได โดย

ภายในของกลองนจะเปนดงรป

-70-

ภาพท 3.7 แสดงการทางานของ State Flow ใน MCU 2

จากภาพท 3.7 จะเหนวาพวกผมแบงการทางานออกเปน 4 โหมดดวยกน

1. Calculating mode

เปนโหมดททาหนาทคานวนคาตางๆทรบมาจากผใชซงคาเหลานจะนาไปใชในโหมดอนตอไป

เราสามารถใชโหมดนไดโดยการปดสวตซ ทกตว

-71-

ภาพท 3.8 แสดงการทางานในโหมด Calculation

2. Normal mode

สาหรบ MCU 2 น โหมดนจะทาหนาทเพยงแคแสดงไฟ LED สขาวขนมาเทานน

ภาพท 3.9 แสดงการทางานในโหมด Normal

3. Manual switch mode

โหมดนจะทาหนาทเกบคา Threshold On, off จากผใช รวมทงแสดงไฟ LED สแดงขนมาดวยเมอ

เครองทางานในโหมดน

-72-

ภาพท 3.10 แสดงการทางานในโหมด Manual switch

จะเหนวาในโหมดนผมไดคงคาคาความถและความกวางของสญญาณไวดวย

4. Tilt sensor mode

โหมดนจะเปนการเกบคา Stim_enable เพอนาไปใชใน MCU 1 ตอไป โดยถาคาทสงไปเปน 1 ก

แสดงวาใหกระตนได แตถาคาทสงไปเปน 0 กใหหยดกระตน และเมอเครองทางานในโหมดนกจะมไฟ

LED สเขยวแสดงขนมา

ภาพท 3.11 แสดงการทางานในโหมด Tilt sensor

-73-

3.2 การพฒนาฮารตแวร

ขนตอนการพฒนาฮารตแวร

3.2.1 ศกษาและเลอกอปกรณทจะใชในการทา

สาหรบอปกรณทอยบน PCB ประกอบดวย

1. จอ LCD

ภาพท 3.12 แสดง LCD ทใช

ภาพท 3.13 แสดง Schematic ของ LCD

-74-

2. MCU 2 ตว

ภาพท 3.14 แสดงบอรด Fio std ทใช

ภาพท 3. 15 แสดงรายละเอยด PIN และ JP บนบอรด

-75-

คณสมบตทางดานฮารตแวรของบอรด

– M3 Processor (STM32F103RET6)

aximum speed

DC

ock)

- USB or External (up to 15VDC) supply input selection via jumper settings A) LDO regulator for internal and external circuitries

3. Accelerometer Sensor

าพ Accelerometer Sensor ทใช

TMARM 32-bits Cortex

90MIPS m64 KBytes SRAM 51 GPIO 16 channels 12-bit, 1μs A5 USART, 3 SPI, 2 I2C, 1 CAN 2 channels 12-bit DAC 496Kbytes available flash memory Two onboard crystals -32 kHz (for real-time cl -8 MHz (for MCU clock) Power supply management

- 3.3V (up to 800m

ภาพท 3.16 แสดงภ

เปน เซนเซอรทเอาไวใชในโหมด Tilt Sensor

-76-

4. สว

งตวหมนปรบคาพารามเตอรตางๆทใช

000 us

s

5. สวตซ ปรบเปลยนการทางานในโหมดตางๆ

มดทใช

โดย สวตซ จะเรยงตามลาดบคอ SW1, SW2, SW3 และ Hold

ตซ ปรบคาความถ ความกวางสญญาณ ชวงกระตน และชวงหยดกระตน

ภาพท 3.17 แสดงภาพขอ

คาความถจะปรบไดตงแต 0 – 120 Hz.

ความกวางสญญาณจะปรบไดทละ 100 us ตงแต 0 – 1

ชวงกระตนและหยดกระตนจะปรบไดตงแต 0 – 60

ภาพท 3.18 แสดงรป Switch เปลยนโห

-77-

6. LED สตางๆ

ดง LED ทแสดงการทางานในโหมดตางๆ

de

ode

ภาพท 3.20 แสดงอปกรณปรบเปลยนแอมพลจด

ภาพท 3.19 แส

สเหลองแสดงวาเครองกระตนอยหรอไม

สขาวแสดงการทางานใน Normal mode

สแดงแสดงการทางานใน Manual switch mo

สเขยวแสดงการทางานใน Tilt sensor m

7. อปกรณปรบเปลยนแอมพลจด

-78-

8. lectrode

ภาพท 3.21 แสดงรป Electrode ทใช

9. ower s

ภาพท 3.22 แสดง Power supply ทใช

E

P upply

-79-

3.2.

แบบ PCB โดยใชโปรแกรม Eagle ซงโปรแกรมทไดหลงจากออกแบบเรยบรอย

แลวเปนดงรป

ภาพท 3.23 แสดง PCB ทออกแบบเรยบรอยแลว

เมอสงทา PCB เรยบรอยและจดวางอปกรณตางๆกจะไดผลลพธทสมบรณดงรป

2 ออกแบบ PCB และจดอปกรณตางๆ

พวกผมออก

-80-

ภาพท 5.24 แสดงภาพรวมของอปกรณทเสรจสมบรณ ( Side view )

ภาพท 5.25 แสดงภาพรวมของอปกรณทเสรจสมบรณ ( Top view )

-81-

บทท 4

การทดลองและผลการทดลอง

4.1 การทดลองการทางานในโหมด Calculation

หลงจากททดลองพบวาเครองสามารถเกบคาพารามเตอรตางๆไดเปนอยางด

ภาพท 4.1 แสดงการทางานในโหมด Calculation

4.2 การทดลองทางานในโหมด Normal

หลงจากทดลองแลวพบวาเครองสามารถกระตนไดตลอดเวลาและไดสญญาณทมคาตรงตาม

ทตงคาไวในโหมดแรกทกประการ

-82-

ภาพท 4.2 แสดงการทางานในโหมด Normal

ภาพท 4.3 แสดงสญญาณทไดเมอตงคาไวคาหนง

4.3 การทดลองการทางานในโหมด Manual switch

หลงจากทดลองแลวพบวาโหมดนสามารถเกบคา Threshold On, off ไดตามปกต และ

สญญาณทไดกออกมาตามทตงคาไว

-83-

ภาพท 4.4 แสดงการทางานในโหมด Manual switch

4.4 การทางานในโหมด Tilt sensor

การทางานในโหมดนเครองสามารถรบคาจากเซนเซอรไดอยางถกตองและสามารถกระตน

ไดตามคา Threshold On, off ทตงเอาไว

-84-

ภาพท 4.5 แสดงการทางานในโหมด Tilt sensor

ภาพท 4.6 แสดงการวดความกวางและความถสญญาณ

-85-

บทท 5

บทวจารณและสรปผล

5.1 ปญหาและการแกไข

ในตอนแรกทพวกผมไดลองใช Simulink สรางโปรแกรมขนมากไดลองอยหลายวธดวยกนและได

พบปญหาตางๆแตกตางกนออกไป ซงพวกผมกไดพยายามแกไปทละจดๆจนทายทสดไดโปรแกรมท จาลอง

ออกมาแลวผลลพธตรงตามความตองการ แตกตองแลกกบการใช MCU ถง 2 ตวดวยกน

จากนนเมอพวกผมพฒนาทางดานซอรฟแวรเรยบรอยแลวจงเรมลงมอทาดานฮารตแวร ในตอนแรก

พวกผมเลอกทจะใชบอรดไขปลาในการสรางเครองกระตนกลามเนอ แตปรากฎวาพอบดกรขาตางๆเรยบรอย

แลว ตอนทดลองผลลพธสายกลบไมแนนหรอชอตตรงนนทตรงนท ทาใหพวกผมตดสนใจทจะกดบอรด

แทน ซงกตองใชเวลาในการออกแบบและสงทาอกหลายวนจนไดสงทตองการออกมา

5.2 แนวทางการพฒนาตอ

ตดตรงขาทอนลางแทนนอง ใชอเลกโทรดแบบอยในสายรด 1.

อาจปรบเปลยนโปรแกรมบางอยาง เพอใหผปวย เดนไดอยางเปนธรรมชาตมากขน 2.

เปลยนสวตซ Hold เปนแบบปมกดเพอใหใชงานไดงายขน 3.

5.3 สรปผลและวจารณ

จากการทดสอบเครองกระตนกลามเนอทสรางขนมาพบวาเครองสามารถกระตนไดตามตองการตาม

แผนทวางไว แตรปลกษณภายนอกของเครองยงไมไดถกทาใหออกมาในรปทสวยงามพรอมใช เนองจาก

เวลาททาโครงการคอนขางจากด อกทงเครองนยงสามารถนาไปพฒนาตอเพอเพมระบบประยกตตางๆท

ผใชตองการไดอกดวย

-86-

บรรณานกรม

www.salisburyfes.com

www.comforttechnologies.com

www.empi.com

www.aimagin.com

http://health.kapook.com/view1703.html

-87-