Heart Valve Disease Classification Using Neural Network
Journal: International Journal of Science and Research (IJSR) (Vol.4, No. 6)Publication Date: 2015-06-05
Authors : Ashish Shelke; V.B. Baru;
Page : 370-372
Keywords : Empirical Mode Decomposition; First heart sound S1; Gaussian distribution; Intrinsic Mode Function; Kurtosis; Second heart sound S2; neural Network; Back propagation algorithm;
Abstract
Heart valve diseases such as aortic stenosis, mitral regurgitation, mitral stenosis and aortic regurgitation can be audible directly by stethoscope. Valve diseases are characterized by systolic murmur and diastolic murmur features. Segment heart sound wave into S1, S2 systolic murmur and diastolic murmur then segments features given to artificial neural network to classify disease. Instead of giving complete PCG wave to ANN give segments and test disease for those segment only then cumulative results of each segment will give final valve disease name. ANN is trained with 49 S1 components, 43 systolic segments consists of 15 of aortic stenosis and 11 of mitral regurgitation.37 diastolic segments used to classify mitral stenosis and aortic regurgitation. These ANN are giving 84.93 % accuracy for S1 test signals having 12 test segments. Systolic gives 88.73 % accuracy and diastolic gives 91.72 % accuracy.
Other Latest Articles
- Antimicrobial Efficacy of Preservatives used in Skin Care Products on Skin Micro Biota
- A Novel Hemagglutinin from Seeds of Dregea volubilis
- A Steady-State Genetic Algorithm for Traveling Salesman Problem with Pickup and Delivery
- Synthesis, Characterization, Biological Activity and Use of Thiourea Derivative ( PhTTU ) Studies of dinitrate-1-phenyl-3-(thiazol-2-yl)-2-thioureato Cd(II) ion Complex
- Design System for Discrimination Patterns Geometric Shapes using Neural Networks (Hamming @ Max net)
Last modified: 2021-06-30 21:49:27