Enhancing Lung Disease Diagnosis via Semi-Supervised Machine Learning

TL;DR

This paper explores semi-supervised machine learning techniques, combining MFCC and CNN models, to improve lung disease detection accuracy while reducing reliance on manual annotations.

Contribution

It introduces semi-supervised modules like Mix Match, Co-Refinement, and Co Refurbishing to enhance detection performance in lung sound analysis.

Findings

Achieved 92.9% accuracy with semi-supervised modules

Improved detection accuracy by 3.8% over baseline

Addresses challenges of limited labeled data and individual differences

Abstract

Lung diseases, including lung cancer and COPD, are significant health concerns globally. Traditional diagnostic methods can be costly, time-consuming, and invasive. This study investigates the use of semi supervised learning methods for lung sound signal detection using a model combination of MFCC+CNN. By introducing semi supervised learning modules such as Mix Match, Co-Refinement, and Co Refurbishing, we aim to enhance the detection performance while reducing dependence on manual annotations. With the add-on semi-supervised modules, the accuracy rate of the MFCC+CNN model is 92.9%, an increase of 3.8% to the baseline model. The research contributes to the field of lung disease sound detection by addressing challenges such as individual differences, feature insufficient labeled data.