Implementation of Real-Time Respiratory Disease Classification with Improved Convolutional Neural Networks (ICNN)

Abstract

Automatic detection of respiratory diseases plays a crucial role in modern healthcare, offering several benefits in terms of efficiency, accuracy, and timely intervention. With the integration of advanced technologies such as machine learning and deep learning, automated systems can analyze respiratory data and swiftly identify various respiratory conditions. This approach eliminates the need for manual analysis, reducing the time required for diagnosis and allowing for prompt medical attention. This paper presents an Automatic Respiratory Data Classification System utilizing an Improved Convolutional Neural Network (CNN) and wavelet transform applied to real-time clinical data. The Respiratory Classification System (RCS) demonstrates robust performance with impressive Normal Respiratory Detection Rates (NRDR) of 98% for normal male data and 95% for normal female data. High accuracy is also achieved in classifying abnormal respiratory data, with NRDRs of 96.67% for abnormal male data and 96% for abnormal female data. The comprehensive evaluation on a substantial dataset results in an outstanding Respiratory Detection Rate (RDR) of 98.8%. The proposed Improved CNN algorithm attains a remarkable RDR with low computational time, showcasing its potential for respiratory disease classification.