17
Cardiovascular wave analysis module for Scilab Serge Steer

ScilabTEC 2015 - Inria

Embed Size (px)

Citation preview

Page 1: ScilabTEC 2015 - Inria

Cardiovascular wave analysis module for Scilab

Serge Steer

Page 2: ScilabTEC 2015 - Inria

The heart : a pump

To o

rgan

sO

2

To the lungs

From the lungsO2

Fro

m o

rgan

sVentricular systole (contraction):● isovolumic contraction

➢ closes the atrioventricular valves ➢ opens sigmoid valves

● ejection

Ventricular diastole (relaxation)● isovolumic relaxation

➢ Opens the atrioventricular valves● filling

➢ Fast➢ Slow➢ Atrial systole

Page 3: ScilabTEC 2015 - Inria

The heart : a regulated pump

Cellular oxygenation, in particular for brain

Long term control : hormones

Short term control : autonomic nervous system

Baroreceptor control loop: two antagonistic actions

- sympathetic = accelerator (fast)

- parasympathetic = inhibitor (slow)

Control also acts on arteries and veins

Sino atrial node

Page 4: ScilabTEC 2015 - Inria

Cardiac electrical activity

The heart contraction is driven by electric depolarization of myocyte cells inducted by cardiac fibers.

Autonomous impulse generator

Cardiac fibers

Autonomic nervoussystem

Page 5: ScilabTEC 2015 - Inria

Trans membrane voltage

DepolarizationSodium ions

Re polarizationPotassium ions Restauration of ionic balance

Cell membranes form electric dipoles

Page 6: ScilabTEC 2015 - Inria

ECG

Electric dipoles generated by cells membranes of the heart produces an electrical potential at the surface of the thorax.

A set of electrodes, typically 3 to 10, are used to measure the electrical potential at different thorax locations.

The electrocardiogram (ECG) records voltage evolution deduced from the measured potentials

Cheap and non invasive

Page 7: ScilabTEC 2015 - Inria

Typical ECG lead trace

P wave Atrial depolarization

QRS complex

Ventricular depolarization

T wave Ventricular repolarization

Page 8: ScilabTEC 2015 - Inria

Applications of the ECG to cardiological diagnosis

● Arrhythmias detections● Disorders in the activation sequences● Increase in wall thickness or size of the atria

and ventricles ● Myocardial ischemia (coronary atherosclerosis)

and infarction ● ...

Heart rate and ECG morphology analysis provides cheap diagnostic elements

Page 9: ScilabTEC 2015 - Inria

The Cardiovascular Wave Analysis module

Designed for – Long term ECG

– Multi-leads records

– Off line analysis

Consists of five classes of tools– Data acquisition

– Pretreatment

– Event detection

– Analysis

– Visualization

Page 10: ScilabTEC 2015 - Inria

Data acquisitionSeveral file formats

● ISHNE Holter (Task Force standard)● TMS32 ( signals recorded using TMS ADC system and PortiLab)● WFDB (WaveForm DataBase used by Physionet)

Can be read and converted into a Scilab structure:

S=readTMS("ECG2.poly5",[25 27]);

viewECG(S(:,1))

Or converted to Scilab specific format (ecgs)

Easy for batch processing

Page 11: ScilabTEC 2015 - Inria

Pretreatment● ECG subsampling● ECG FIR filtering ● Power line interference removal

● ECG detrending● Finding and removing outliers

viewECG([S(:,1) ECGSubstractPLI(S(:,1),50)])

Page 12: ScilabTEC 2015 - Inria

Detections

● R peaks heart rate● T wave ends Q-T segment● All events

S=extractPartFromEcgsFile(“P5J0.ecgs",... 100,3000);S1=ECGDetrend(S);S1=ECGSubstractPLI(S1,50); L=ECGDetections(SynthesisECG(S1)); ECGShowDetections(S(:,1),L)

Page 13: ScilabTEC 2015 - Inria

Analysis tools

Classical signal processing methods (heart rate spectral analysis)

Multichannel non stationary signal analysis

Baroreflex analysis (Arterial blood pressure effect on heart rate)

Complex demodulation (Breathing rate effect on heart rate)

Time frequency analysis

Time domain characteristics

Interactive toolsHRVAS (Heart Rate Variability Analysis System),

TimeFrequencyTool

Page 14: ScilabTEC 2015 - Inria
Page 15: ScilabTEC 2015 - Inria

The contribution of Scilab

● Signal processing and analysis● Data file handling● Graphics● Graphical user interface● Community● Free multi-plateforms software

Page 16: ScilabTEC 2015 - Inria

Acknowledgments- Detection algorithms :

● Quighua Zhang (INRIA)

- Heart rate analysis algorithms: ● Alexandro Monti (INRIA)● John T. Ramshur (U. Menphis)

- Signal processing● D.E. Lake, J.R. Moorman and C. Hanqing● Wavelets :Christopher Torrence (Exelis), Gilbert. P. Compo (U. Colorado)● George B. Moody, Cambridge● Time Frequency toolbox: François Auger (CNRS), Holger Nahrstaedt (TU Berlin)

- Data● Physionet● François Cottin (U. Paris Sud)● Lisa Guigue ( INRIA)

- Testing● Claire Medigue (INRIA)● Lisa Guigue (INRIA)

- Atoms Module (http://atoms.scilab.org/toolboxes/CWA)● Dominique Callens (Scilab-Enterprises)

Page 17: ScilabTEC 2015 - Inria

The end

Thanks