الفهرس | Only 14 pages are availabe for public view |
Abstract Listening and interpreting heart sounds by a stethoscope is used as an important component of diagnosing the cardiac diseases. However this practice has always been vulnerable to poor audibility, inter-observer variations (between different physicians). Thus computerized analysis of heart sounds for objective diagnosis is seen as a probable aid. In the present research the normal and abnormal heart sounds signals were analyzed to extract their specifications by the multifractal detrended fluctuation analysis (MFDFA) and the support vector machine (SVM) to distinguish between the normal heart sound and the abnormal one using the MATLAB. Heart sounds and murmurs are of relatively low intensity and are band limited to about 10–1000 Hz. The cardiac cycle has 4 components S1 (100-160 ms and 10-140 HZ), S2 (80-140 ms and 10-400 HZ), S3 (40-80 ms and 15- 60 HZ) and S4 (30-60 ms and 15-45 HZ). The present research system is very robust, computationally simple and yielded to accuracy of 96.875%. The algorithm makes analysis to the heart sounds signals, extracts the specification and finally differentiates between the normal and abnormal heart sounds signals. |