RESPIRATORY SOUND BASED DISEASE DETECTION SYSTEM USING MACHINE LEARNING

Computer Engineering

Based on sounds that people make, like their voice, dry cough, and breath, the models that are based on Machine Learning (ML) were used in the real world to identify a number of diseases. Using the CNN (Convolutional Neural Network), AI-based machines can solve numerous real-world issues. A modified Convolutional Neural Network (DCNN) that will outperform previous models has been proposed and implemented for automatic disease diagnosis based on human respiratory sounds like voice, dry cough, and breath. Human sounds fall into two main categories: speech sounds and non-speech sounds. The proposed methodology takes into account the non-speech human sounds of laughter, scream, sneeze, and snore. Coughing is a common symptom of many diseases. It is helpful to take into consideration both the type of cough and the sound it makes when diagnosing a disease. We demonstrated that AI algorithms can identify and diagnose COVID19, pertussis, pneumonia, pulmonary enema, asthma, and tuberculosis (TB) using the cough symptom. Additionally, we discovered the various approaches that yielded the most accurate diagnoses of respiratory conditions using cough samples. Consequently, we develop a machine learning architecture for the proposed system's disease classification based on human sounds..

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.