Advancements in Obstructive Sleep Apnea Diagnosis and Screening Through Artificial Intelligence: A Systematic Review

Area of Science:

• Medical Informatics
• Biomedical Engineering
• Artificial Intelligence in Medicine

Background:

• Obstructive sleep apnea (OSA) is a widespread condition with significant healthcare costs, often underdiagnosed due to the inaccessibility of traditional polysomnography (PSG).
• Artificial Intelligence (AI), encompassing Machine Learning (ML) and deep learning (DL), presents innovative solutions for OSA screening and diagnosis.
• This systematic review focuses on AI-driven models for OSA detection over the past decade.

Purpose of the Study:

• To systematically review and evaluate the efficacy of AI-powered models in screening and diagnosing obstructive sleep apnea (OSA).
• To assess the potential of AI in improving the accessibility and accuracy of OSA diagnosis compared to conventional methods.

Main Methods:

• A comprehensive literature search was conducted across PubMed/MEDLINE, Google Scholar, and SCOPUS databases.
• Included studies were original English articles detailing the use of ML algorithms for OSA diagnosis and prediction in suspected patients.
• The systematic review is registered with PROSPERO (Registration ID: CRD42024563059), with the last search in June 2024.

Main Results:

• Sixty-five studies, encompassing data from 109,046 patients, met the inclusion criteria.
• AI algorithms demonstrated superior performance in OSA detection, often surpassing traditional tools in accuracy, sensitivity, and specificity.
• Anthropometric indexes, particularly within logistic regression models, were the most frequently utilized features.

Conclusions:

• AI applications hold significant potential for enhancing OSA diagnosis and screening, leading to improved patient outcomes and reduced healthcare burdens.
• Standardization of datasets and rigorous validation are crucial for the widespread adoption and reliability of AI tools in OSA management.