ENHANCED MACHINE LEARNING FOR SPIROMETRY CURVE ASSESSMENT IN CHRONIC RESPIRATORY DISEASES

Abstract

Chronic respiratory diseases, including Chronic Obstructive Pulmonary Disease (COPD) and asthma, pose a significant global health challenge with far-reaching economic implications. This study introduces a machine learning-based approach for predicting exacerbation risks in respiratory dis- eases, specifically focusing on COPD and asthma. The frame- work employs an advanced machine learning architecture to enable real-time and precise detection of respiratory events. spirometry standards mandate that correct maneuvers shouldbedevoidofcoughartifacts,especiallyintheinitialseconds of forced exhalation. Achieving this standard becomes inherently challengingwhenpatientsmanifestincreasedcoughingtendencies duringtheexamination.Thisstudyaimtofacilitateunsupervised or minimally supervised spirometry measurements, expanding accessibility in diverse settings, including home monitoring and general practitioner oversight. By prioritizing specificity in the machinelearningtechniques,themodelachievesacomprehensive understanding and accurate classification of symptoms, thereby contributing to personalized risk assessment.