Abstract
This article presents a systematic literature review examining the current applications of artificial intelligence (AI) in medical diagnosis, their documented clinical outcomes, and perspectives for implementation in low- and middle-income country healthcare systems. Publications from 2015 to 2024 indexed in PubMed, Scopus, Web of Science, and Cochrane Library were systematically analysed. The review covers AI diagnostic applications across imaging-based specialties including radiology, pathology, and ophthalmology, as well as clinical decision support in internal medicine, and addresses methodological quality, safety considerations, and implementation challenges. The analysis demonstrates that AI diagnostic systems achieve performance comparable or superior to specialist clinicians in specific, well-defined diagnostic tasks — including diabetic retinopathy screening (AUC 0.99), pneumonia detection on chest radiographs (AUC 0.93), and melanoma identification (AUC 0.91) — while facing significant barriers to real-world clinical deployment including training data quality, algorithmic bias, and regulatory uncertainty. Implications for AI diagnostic implementation in Uzbekistan's healthcare system are identified.
References
This work is licensed under a Creative Commons Attribution 4.0 International License.