Performance analysis of deep learning algorithms for classifying chronic obstructive pulmonary disease

Abstract

Nowadays, Deep learning (DL) and machine learning (ML) play a vital role in furnishing solutions to the medical problems. Owing to their accurate and timely forecasting models and results, ML and DL algorithms are being embraced by the medical professionals for early detection and prompt treatment of different diseases. The respiratory diseases like Chronic Obstructive Pulmonary Disease (COPD) are emerging and need an early diagnosis. The major methods for diagnosing COPD involve expensive and unsuitable spirometer and imaging equipment. In this paper, an analysis of cough sound of the patients and identification of COPD severity levels using ML and DL algorithms has been reported. The study includes experiments conducted using Librosa library and used CNN, RNN, LSTM, and MLP algorithms for detecting COPD severity levels.

URN:NBN:sciencein.jist.2024.v12.745