Phonocardiogram Classification

Aritra Roy Gosthipaty

Aritra Roy Gosthipaty

Kolkata, West Bengal

Classification of normal and abnormal sound based on the phonocardiogram(hear sounds) using classic machine learning algorithms. ...learn more

Project status: Under Development

Artificial Intelligence

Intel Technologies
Intel Python

Code Samples [1]

Overview / Usage

Phonocardiogram is a chart or record of the sounds made by the heart. The aim of the project is to classify a heart as normal and abnormal based on the phonocardiogram of the heart. This will open up a wide spectre of innovation and research in the biomedical field. This will help people get treated properly and quicker in a very cost effective manner.

Methodology / Approach

The approach towards the project is divided into parts.

  1. Segmentation: The pre requisite of the project is segmentation of the phonocardiogram. The segmentation is done with Hidden Markov model which lets segment a signal into 4 parts namely s1,s2,s3 and s4. The segmentation helps usconcentrate on the murmurs(noises) in each section and then extract viable features for the classification task.
  2. Classification: A perceptron model was proposed that would be based on statistical features of a signal. Support Vector machine was proposed too. Both the apporaches were no where near to make a good precision or accuracy.

Technologies Used

  • Intel Python
  • Matlab

Repository

https://github.com/forkbomb-666/phonocardiogram

Collaborators

There are no people to show.

Comments (0)