AI-oximetry models show 91.1% sensitivity and 88.4% specificity for obstructive sleep apnea diagnosisAI models show high accuracy in diagnosing obstructive sleep apnea

BACKGROUND: Obstructive sleep apnea (OSA) affects 38% of the population, yet over 90% of cases remain undiagnosed. The gold standard for diagnosis, polysomnography, requires specialized equipment and trained personnel, making it inaccessible in primary care and acute settings. With artificial intelligence (AI) advancements, oximetry-based AI models have emerged as potential alternatives for OSA diagnosis. OBJECTIVE: This meta-analysis aims to evaluate the diagnostic accuracy of AI models trained on pulse oximetry readings in diagnosing OSA. METHODS: A systematic search was conducted across Medline/PubMed, Embase, Scopus, Web of Science, and IEEE Xplore databases from inception to January 3, 2026. Studies that evaluated the diagnostic accuracy of AI models trained on oxygen saturation recordings, compared to the apnea-hypopnea index (AHI) as the reference standard, were included and screened by 2 blinded independent reviewers. Models were evaluated using Bayesian bivariate meta-analysis and meta-regression. Publication bias was examined using a selection model approach, while risk of bias and evidence quality were assessed with Quality Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2) and Grading of Recommendations Assessment, Development, and Evaluation (GRADE). RESULTS: From 13,986 screened articles, 25 studies met the inclusion criteria, encompassing 23,171 participants with a mean age of 40 (SD 10.6) to 63 (SD 13.3) years and a BMI of 25 to 37 kg/m2. AI-oximetry models demonstrated a pooled sensitivity of 91.1% (95% credible interval [CrI] 89.7%-92.4%) and specificity of 88.4% (95% CrI 85.3%-90.8%). Neural network classifiers achieved the highest sensitivity (92.7%) and specificity (91.3%). Deep learning feature extraction models were significantly higher in sensitivity (by 3.7%; 95% CrI 0.9%-6.9%) than domain expert-based approaches. Sensitivity decreased slightly with higher AHI cutoffs, while specificity increased by 16.6% from an AHI cutoff of ≥5 to ≥30. Sensitivity analyses showed that even with up to 40% probability of an unpublished study, changes in accuracy were modest (area under the curve: 0.902 to 0.877). QUADAS-2 and GRADE assessments found low-moderate risk of bias with high overall quality of evidence. CONCLUSIONS: AI-oximetry models showed high diagnostic accuracy for OSA across models and AHI cutoffs, performing better than or comparably to traditional overnight oximetry and home sleep apnea tests. This review provides the first pooled quantitative synthesis of AI models trained solely on oximetry data, with additional evaluations of publication bias and methodological limitations. Prior reviews were largely narrative or used alternative AI inputs other than oximetry. This study advances the field by offering a clearer and more reliable evidence base on pooled AI oximetry performance. These findings support the potential of oximetry-based AI as a convenient and scalable tool for OSA screening and diagnosis, with potential real-world applications in both primary care and inpatient settings for early identification of high-risk patients. Prospective external validation in diverse populations and low-prevalence settings is still needed before widespread real-world use.