Sucharita Mitra

A Rough-Set-Based Inference Engine for ECG Classification

In this paper, a rule-based rough-set decision system for the development of a disease inference engine is described. For this purpose, an offline-data-acquisition system of paper electrocardiogram (ECG) records is developed using image-processing techniques. The ECG signals may be corrupted with six types of noise. Therefore, at first, the extracted signals are fed for noise removal. A QRS detector is also developed for the detection of R-R interval of ECG waves. After the detection of this R-R interval, the P and T waves are detected based on a syntactic approach. The isoelectric-level detection and base-line correction are also implemented for accurate computation of different attributes of P, QRS, and T waves. A knowledge base is developed from different medical books and feedbacks of reputed cardiologists regarding ECG interpretation and essential time-domain features of the ECG signal. Finally, a rule-based rough-set decision system is generated for the development of an inference engine for disease identification from these time-domain features

A lossless ECG data compression technique using ASCII character encoding

A software based lossless ECG compression algorithm is developed here. The algorithm is written in the C-platform. The algorithm has applied to various ECG data of all the 12 leads taken from PTB diagnostic ECG database (PTB-DB). Here, a difference array has been generated from the corresponding input ECG data and this is multiplied by a large number to convert the number of arrays into integers. Then those integers are grouped in both forward and reverse direction, out of which few are treated differently. Grouping has been done in such a way that every grouped number resides under valid ASCII value. Then all the grouped numbers along with sign bit and other necessary information are converted into their corresponding ASCII characters. The reconstruction algorithm has also been developed in using the reversed logic and it has been observed that data is reconstructed with almost negligible difference as compared with the original (PRD 0.023%).

ECG signal compression using ASCII character encoding and transmission via SMS

Abstract Software based efficient and reliable ECG data compression and transmission scheme is proposed here. The algorithm has been applied to various ECG data of all the 12 leads taken from PTB diagnostic ECG database (PTB-DB). First of all, R-peaks are detected by differentiation and squaring technique and QRS regions are located. To achieve a strict lossless compression in the QRS regions and a tolerable lossy compression in rest of the signal, two different compression algorithms have used. The whole compression scheme is such that the compressed file contains only ASCII characters. These characters are transmitted using internet based Short Message Service (SMS) and at the receiving end, original ECG signal is brought back using just the reverse logic of compression. It is observed that the proposed algorithm can reduce the file size significantly (compression ratio: 22.47) preserving ECG signal morphology.

An automated data extraction system from 12 lead ECG images

A software based normalized ECG data acquisition system is developed for both normal and abnormal ECG records. This system can transfer wave data recorded on paper to the digital time database. A flatbed scanner is used to form an image database of each 12 lead ECG signal. These TIF formatted gray tone images are then converted into two tone binary images with the help of histogram analysis. Smearing runlength technique is used to remove the vertical and horizontal line segments of graphical papers. Thinning algorithm is applied to each image to obtain the skeleton (1 pixel representation) of each image, which is essential to avoid excess data points in the database. After extracting pixel to pixel co-ordinate information of images of each of the signal of 12 lead ECG records, the data are sorted to regenerate the signal. From standard deviation of the database a graphical analysis is performed to examine the consistency of our database.

An ECG data compression method via R-Peak detection and ASCII Character Encoding

Efficient and reliable electrocardiogram (ECG) compression system can increase the processing speed of real-time ECG transmission as well as reduce the amount of data storage in long-term ECG recording. In the present paper, a software based effective ECG data compression algorithm is proposed. The whole algorithm is written in C- platform. The algorithm is tested on various ECG data of all the 12 leads taken from PTB Diagnostic ECG Database (PTB-DB). In this compression methodology, all the R-Peaks are detected at first by differentiation technique and QRS regions are located. To achieve a strict lossless compression in QRS regions and a tolerable lossy compression in rest of the signal, two different compression algorithms have developed. In lossless compression method a difference array has been generated from the corresponding input ECG “Voltage” values and then those are multiplied by a considerably large integer number to convert them into integer. In the next step, theses integer numbers are grouped in both forward and reverse direction maintaining some logical criteria. Then all the grouped numbers along with sign bit and other necessary information (position of critical numbers, forward/reverse grouping etc.) are converted into their corresponding ASCII characters. Whereas in lossy area, first of all, the sampling frequency of the original ECG signal is reduced to one half and then, only the “Voltage” values are gathered from the corresponding input ECG data and those are amplified and grouped only in forward direction. Then all the grouped numbers along with sign bit and other necessary information are converted into their corresponding ASCII characters. It is observed that this proposed algorithm can reduce the file size significantly. The data reconstruction algorithm has also been developed using the reversed logic and it is seen that data is reconstructed preserving the significant ECG signal morphology.

ECG feature extraction using differentiation, Hilbert transform, variable threshold and slope reversal approach

An accurate and reliable ECG feature extraction algorithm is presented in this paper. ECG samples are de-noised and its first derivative and Hilbert transform are computed. Sample having maximum amplitude in the transformed domain is found out and those samples having amplitudes within a lead wise specified threshold of that maximum are marked. In the original signal, where these marked samples undergo slope reversals are spotted as R-peak. On the left and right side of the R-peak, slope reversals are identified as Q and S peak, respectively. QRS onset-offset points, T and P waves are also detected. ECG baseline modulation correction is done after detecting characteristics points. The algorithm offers a good level of Sensitivity, Positive Predictivity and accuracy of R peak detection. Each wave and segment duration and each peak height is measured. Measurement errors of extracted ECG features are calculated. The algorithm is implemented on MATLAB 7.1 environment.

A combined application of lossless and lossy compression in ECG processing and transmission via GSM-based SMS

Abstract This paper presents a software-based scheme for reliable and robust Electrocardiogram (ECG) data compression and its efficient transmission using Second Generation (2G) Global System for Mobile Communication (GSM) based Short Message Service (SMS). To achieve a firm lossless compression in high standard deviating QRS complex regions and an acceptable lossy compression in the rest of the signal, two different algorithms have been used. The combined compression module is such that it outputs only American Standard Code for Information Interchange (ASCII) characters and, hence, SMS service is found to be most suitable for transmitting the compressed signal. At the receiving end, the ECG signal is reconstructed using just the reverse algorithm. The module has been tested to all the 12 leads of different types of ECG signals (healthy and abnormal) collected from the PTB Diagnostic ECG Database. The compression algorithm achieves an average compression ratio of ∼22.51, without any major alteration of clinical morphology.

Noise reduction and ECG feature extraction using interpolation and Hilbert transform

An effective and reliable noise reduction and Electrocardiogram (ECG) feature extraction algorithm is proposed in this paper. Contaminated ECG samples are de-noised using a Butterworth lowpass and IIR notch filter. First derivative using Lagrange Five Point Interpolation formula and Hilbert Transform of those ECG samples are computed. Sample having maximum amplitude is found out from the transformed data and those samples having amplitude within a lead wise specific threshold of that maximum are selected. The point where those selected samples undergo slope alteration in the original time domain ECG signal is marked as R peak. After successful identification of R peak points, base line modulation correction is implemented using an empirically determined formula. Q and S points are identified by finding minimum amplitude on the either side of the most recently detected R peak. QRS onset and offset points are also detected. After detecting all these characteristic points, Heart Rate, R, Q and S peak heights and QRS duration are measured. Errors in these extracted ECG features are also calculated. The algorithm offers a good level of Sensitivity (99.84%), Positive Predictivity (99.84%) and Detection Accuracy (99.84%) of R peak. Different types of ECG data of all the 12 leads taken from PTB diagnostic ECG database (PTB-DB) is used for testing the performance of the proposed module.

A robust technique for delineation and features extraction of ECG signal from mobile-phone photography

This paper reports the development of a software suite to be accessed in future with any General Packet Radio Service (GPRS) or High Speed Packet Access (HSPA) enabled mobile phone or Personal Digital Assistant (PDA) for the extraction and analysis of disease-related features from the photograph of paper based ECG records. In India and other developing countries, the cheaper paper based ECG machines are prevalently used. In rural areas of these countries cardiac diseases are still the major silent killers due to the acute dearth of qualified cardiologists. One way of addressing this problem is Tele-medicine which necessitates an intelligent cardiac parameter extraction algorithm. In our bid to address this requirement, an algorithm is developed with the help of few image processing techniques. Initially, the background noises i.e. the gridlines are removed by thresholding technique. Applying the Sauvola method for adaptive image binarization and subsequent morphological operations to get pure ECG signature on white background, this algorithm intelligently applies Bresenhams line drawing algorithm to join the disjoined ECG signature where required. Then thinning has been done for extraction of digital time-plane data and then Discrete Wavelet Transform (DWT) and water reservoir based pattern recognition technique are subsequently used to delineate other important time-plane features for ECG interpretation.

A novel approach for delineation and feature extraction in QRS complex of ECG signal

This paper reports the development of a software suite to be accessed in future with any General Packet Radio Service (GPRS) enabled mobile phone or Personal Digital Assistant (PDA) for the extraction and analysis of disease-related features from the images of paper based ECG records. In India and other developing countries, the cheaper paper based ECG machines are mainly used. Also in rural areas of these countries - where there is acute dearth of qualified cardiologists, the cardiac diseases are still the major silent killers. One way of addressing this problem is cellular medicine which necessitates an intelligent cardiac parameter extraction algorithm. In our bid to address exactly this requirement, a software is developed with the help of few image processing techniques. At first, the background noises i.e. the gridlines are removed by thresholding technique. Binarization and thinning has been done for extraction of digital time data and then Discrete Wavelet Transform (DWT) and water reservoir based pattern recognition technique are subsequently used to delineate other important ECG parameters for ECG interpretation.