Enhancing Respiratory Disease Diagnosis through FMCW Radar and Machine Learning Techniques

Ariana Tulus Purnomo (Sampoerna University, Indonesia)
Raffy Frandito (Sampoerna University, Indonesia)
Edrick Hansel Limantoro (Sampoerna University, Indonesia)
Rafie Djajasoepena (Sampoerna University, Indonesia)
Muhammad Agni Catur Bhakti (Sampoerna University, Indonesia)
Ding-Bing Lin (National Taiwan University of Science and Technology, Taiwan)

G-Tech : Jurnal Teknologi Terapan
Vol 8 No 1 (2024): G-Tech, Vol. 8 No. 1 Januari 2024

This study addresses the urgent need for early diagnosis and continuous monitoring of respiratory diseases such as asthma, Chronic Obstructive Pulmonary Disease (COPD), and infectious diseases. We have created a system that combines state-of-the-art machine learning algorithms with frequency-modulated continuous wave (FMCW) radar technology. FMCW radar technology is sensitive to small respiratory movements, allowing real-time monitoring without physical contact. Machine learning algorithms, including Decision Trees, Random Forest, Naïve Bayes, Gradient Boosting, and Support Vector Machines, are employed to classify these waveforms. The Random Forest classifier achieved the highest accuracy score of 94.6%, with Naïve Bayes exhibiting the shortest processing time at 0.055 seconds. We explored various cross-validation methods such as Shuffle Split, K-fold, and Stratified K-fold, with the Shuffle Split method performing best overall in terms of accuracy and time. Our study introduces an integrated system that could revolutionize the early detection, response, and tracking of respiratory diseases and emergencies over time.