an improved method of measurement of ecg parameters for ... - ICMCC

icmcc.org

an improved method of measurement of ecg parameters for ... - ICMCC

AN IMPROVED METHOD OF MEASUREMENT

OF ECG PARAMETERS FOR ONLINE

MEDICAL DIAGNOSIS

H.R SINGH, RAHUL SHARMA,

NITIN SAHGAL, POONAM SETHI,

AHUL KUSHWAH & PRANAV KACHHAWA


ABSTRACT

The accuracy in the online measurement of ECG Parameters has

a decisive role in the better diagnosis and effective treatment of the

diseases.

The present paper describes a Lab-VIEW based programming

using Pan Tompkins method to extract out QRS complex whereas QT

interval measurements were carried out using Mat-lab based Math-

Script module.

Hilbert Transform has been applied on the ECG signal to convert it

into an analytical signal for better peak detection.

Peak detection and other parameters like RR interval, HR and

several time domain measures of Heart Rate Variability such as RR

mean and standard deviations, HR mean and standard deviations,

RMSSSD, NN50 count, pNN50 etc were calculated for several other

clinical applications apart from online disease diagnosis.


ECG PARAMETERS AND THEIR IMPORTANCE

IN DIAGNOSIS.

P wave is produced by muscle contraction of atria.

The shape and duration of P wave indicate atrial

enlargement.

R wave marks the ending of the atrial contraction

and the beginning of ventricular contraction.

Magnitude normally varies from 0.1mV-1.5mV

Narrow and high R wave indicates a physically

strong heart.

T wave marks the ending of ventricular

contraction.

A normal T wave is slight round and symmetrical.

Pointed T wave is a cause of concern.

Tall T wave indicates a certain disease.


Placement of electrodes

Time it takes the impulse to travel from

atria to AV node (Atrio-ventricular conduction

time).

The PR interval

Measured from the onset of the P wave to

onset of the QRS complex

No more than 5 small squares in duration

(0.20 sec)

Prolonged PR interval >0.20 secs. in 1st

degree heart block.


The QRS complex


Represents ventricular contraction.

Measured from the onset of Q wave to the

end of S wave.

Between 0.08 and 0.12 secs in duration (3

small squares).

Since the ventricles contain greater

muscle mass than the atria, the QRS complex

is larger than the P wave.

First negative component of the QRS

complex

Should be less than 25% of the associated

R wave.


METHODOLOGY

QRS and QT detection were performed using Pan Tompkins

while most of the other parameters such as RR interval, BPM,

peaks, RR measurements and several other statistical and

geometric measures were detected using Hilbert Transform.

In order to isolate the QRS energy predominant portion, the

raw signal after acquisition is pre- processed by proper signal

conditioning and removing base line wandering.

These parameters after extracting out from the enhanced

ECG signal have been significantly used in diagnostic

applications with accuracy acceptable to provide higher order

of care to the patient.


PAN-TOMPKINS ALGORITHM

Identifies QRS complex based on the analysis of the slope,

amplitude and width of the QRS.

Bandpass filter formed using low pass and high pass filters

reduce noise in the ECG signal.

It also removes baseline drift.

The differentiator distinguishes QRS complexes from low

frequency P and T waves.

Squaring emphasizes the higher values expected due to QRS

complex and suppresses smaller values related to P and T

waves.


Moving window integrator is required due to the presence of

multiple peaks within the duration of a single QRS, this takes an

average of N samples.

Its output can be used to detect QRS complexes, RR interval

and determine the duration of the QRS complex.


QRS DETECTION

A block diagram implementing the above stated stages in Lab-VIEW

program is shown in figure for QRS detection.

BLOCK DIAGRAM OF QRS DETECTION


INPUT/OUTPUT WAVEFORMS OF

A NORMAL SINUS RHYTHM


QT DETECTION

The gap interval between Q onset

and T offset

values is calculated on

a signal as acquired and processed in the previous section of QRS

complex determination.

The block diagram developed for the detection of QT interval and

Q point and T point time measurements is shown in figure.

BLOCK DIAGRAM FOR DETECTION OF QT INTERVAL


The corresponding QT interval measurement displaying

the QT off

and QT on

values separately along with the signal

acquired is shown below :-

FRONT PANEL OF QT DETECTION


The derivative of the signal acquired after filtering and base line wandering

elimination using median filter is taken on which Hilbert Transform is

performed.

The reason Hilbert Transform is used for these measurements is to turn the

ECG signal to an analytic signal, which gives a better peak to detect.

This process will give an enhanced ECG signal as compared to raw ECG

signal first obtained.

Peak detection.vi algorithm is developed to find all the peaks and their

locations by setting appropriate threshold parameters.

The RR intervals are extracted by measuring the time interval between two

consecutive peaks.

The heart rate is calculated as follows :

Sampling rate *60/ RR interval (ms)


HRV ANALYSIS METHODS

There are different methods of HRV analysis. one of this method

is Time Domain Analysis. This method extracts a few special

measures using only the temporal RR interval signals.

For Time Series Analysis, Time Domain measures are commonly

used. Many measures can be extracted from the original RR interval

signals to show the changes in the ans.

Variables Units Description Statistical Measures

RR Mean & Std S Mean and standard deviation of all RR intervals

HR Mean & Std 1/min Mean and standard deviation of all heart rates.

RMSSD MS Square root of the mean of the sum of squares of

differences between adjacent RR intervals.

NN50 count

Number of pairs of adjacent RR intervals differing

by more than 50 ms in all measurements

pNN50 % NN50 count divided by the total number of all RR

intervals

Geometrical measures

HRV triangular

index

Total number of all RR intervals divided by the

height of the histogram of all RR intervals.


HEART RATE VARIABILITY ANALYSIS

The Heart Rate Variability analysis can be performed in many ways but the

commonly used method is Time Domain Analysis where only the temporal RR

interval signals are used to extract out a few special measures such as mean

and standard deviation (RR mean) of all RR intervals and heart rate (HR),

RMSSD (square root of mean of sum of square of differences between

adjacent RR intervals), NN50 count (Number of pairs of adjacent RR intervals

differing by more than 50 ms) and HRV triangular index (Total no. of all RR

intervals divided by height of the histogram of all RR intervals) etc.


The process of acquiring various signal parameters as stated above.

Acquire raw ECG

signal

Baseline drift

elimination

Hilbert’s

transform

Time domain

HRV analysis

Heart rate

Extract R peaks &

RR intervals

Peak Detection

PROCESS OF ACQUIRING SIGNALPARA METERS


BLOCK DIAGRAM OF TIME DOMAIN HRV ANALYSIS


Start

Data Acquisition

Signal Conditioning

Processing

Display raw ECG waveform, enhanced ECG waveform,

Heart Rate, time domain parameters of HRV

Transmission of data parameters and waveforms

no

Is Heart Rate out

of range?

Warning indicator at remote PC

Do you want to

save the data?

no

Press the save button

Do you want to

quit?

Stop


Figure : ECG data of patient with normal heart rate on Local PC(top) and Remote PC(bottom).


Figure : ECG data of patient with abnormal heart rate on Local PC(top)and Remote PC(below).


CONCLUSION

VIs for the determination of QRS duration using Pan Tompkins

algorithm is developed and stepwise execution of every stage is

displayed in the front panel diagram.

For QT interval measurement, Math-Script tool in Mat-lab is used

and the corresponding QT intervals along with their Q point and T

point time measurements are displayed in the respective front panel

diagram.


A single window VI for the measurement of all time

domain statistical and geometric measures such as RR

interval, HR, HRV time variant analysis, RR and HR mean and

standard deviations, RMSSSD, NN50 count is also developed

and results are displayed.

These results and their interdependencies on each other

are highly desirable for several clinical applications apart from

the online medical diagnosis.

More magazines by this user
Similar magazines