TriSpectraKAN: A novel approach for COPD Detection via Lung Sound Analysis

Abstract This study aims to create an automated, accessible, andcost-effective diagnostic tool for Chronic Obstructive Pulmonary Disease(COPD). Traditional diagnostic methods are expensive, time-consuming,and require specialized equipment. The proposed TriSpectraKAN modelleverages audio-based lung sound features to improve early diagnosis.TriSpectraKAN is a hybrid model combining spectral features and theKolmogorov-Arnold Network (KAN) to analyze lung sounds using Mel-frequency cepstral coefficients (MFCCs), chromagram, and Mel spectro-grams. Each sub-model focuses on a different audio feature, capturingunique sonic signatures. These features are merged through a hybridnetwork for comprehensive analysis. The model, trained on a COPDdataset, was deployed on a Raspberry Pi for real-time use. TriSpec-traKAN achieved 93% accuracy, an F1 score of 0.98, precision of 0.97,and recall of 0.98. This multimodal approach captured a broad rangeof lung sound features, improving diagnosis accuracy compared to tra-ditional methods. The integration of multiple audio features in TriSpec-traKAN enhances COPD diagnosis, demonstrating the potential of AIand machine learning to transform respiratory disease diagnosis throughaccessible tools.