Enhancing Lung Disease Diagnosis via Semi-Supervised Machine Learning

2025-07-20

AI Summary

To address the challenges of scarce manually annotated data and high inter-subject variability in diagnosing pulmonary diseases (e.g., lung cancer, COPD), which severely limit model generalizability, this paper proposes a semi-supervised lung sound analysis framework. Building upon MFCC feature extraction and a CNN-based classifier, the method innovatively integrates three complementary semi-supervised modules—MixMatch, Co-Refinement, and Co-Refurbishing—to jointly optimize training using limited labeled and abundant unlabeled lung sound recordings. Experimental results demonstrate that the proposed approach achieves 92.9% classification accuracy under constrained labeling budgets, outperforming the supervised baseline by 3.8 percentage points. It significantly enhances model robustness and clinical applicability, offering a scalable technical pathway for intelligent auscultation in low-resource healthcare settings.

Technology Category

• Machine Learning: Semi-Supervised Learning
• Computer Vision: Multi-modal Vision
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Web Mining and Content Analysis: Large pretrained models with web data
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization

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.

Problem

• Improving lung disease diagnosis using semi-supervised machine learning
• Reducing reliance on costly invasive traditional diagnostic methods
• Addressing insufficient labeled data and individual differences in detection

Innovation

• Semi-supervised learning for lung sound detection
• MFCC+CNN model with Mix Match modules
• Improved accuracy to 92.9% with reduced annotations