Enhancing artificial intelligence-driven sleep apnea diagnosis: The critical importance of input signal proficiency with a focus on mandibular jaw movements.

Abstract

This review aims to highlight the pivotal role of the mandibular jaw movement (MJM) signal in advancing artificial intelligence (AI)-powered technologies for diagnosing obstructive sleep apnea (OSA). A scoping review was conducted to evaluate various aspects of the MJM signal and their contribution to improving signal proficiency for users. The comprehensive literature analysis is structured into four key sections, each addressing factors essential to signal proficiency. These factors include (1) the comprehensiveness of research, development, and application of MJM-based technology; (2) the physiological significance of the MJM signal for various clinical tasks; (3) the technical transparency; and (4) the interpretability of the MJM signal. Comparisons with the photoplethysmography (PPG) signal are made where applicable. Proficiency in biosignal interpretation is essential for the success of AI-driven diagnostic tools and for maximizing the clinical benefits through enhanced physiological insight. Through rigorous research ensuring an enhanced understanding of the signal and its extensive validation, the MJM signal sets a new benchmark for the development of AI-driven diagnostic solutions in OSA diagnosis.