TRANSFER LEARNING BASED DIAGNOSIS AND ANALYSIS OF LUNG SOUND ABERRATIONS

With the development of computer -systems that can collect and analyze enormous volumes data, medical profession is establishing several non-invasive tools. This work attempts to develop a technique for identifying respiratory sounds acquired by stethoscope voice recording software via machine learning techniques. study suggests trained proven CNN-based approach categorizing sounds. A visual representation each audio sample constructed, allowing resource identification classification using methods like those used effectively describe visuals. We called MelFrequency Cepstral Coefficients (MFCCs). Here, features are retrieved categorized VGG16 (transfer learning) prediction accomplished 5-fold cross-validation. Employing various data splitting techniques, Respiratory Sound Database obtained cutting- edge results, including accuracy 95%, precision 88%, recall score 86%, F1 81%. The ICBHI dataset train test model.