LLM Diagnostic Reasoning in Fetal CNS MRI — ChatGPT, Gemini, and DeepSeek Compared
COMPARISONA retrospective single-center diagnostic accuracy study evaluated the performance of ChatGPT, Gemini, and DeepSeek on text-based diagnostic reasoning for fetal central nervous system (CNS) MRI cases. Fetal MRI plays an important role in evaluating prenatal CNS abnormalities, but expert interpretation and counseling remain challenging — creating a potential role for LLM-based decision support. The study is notable as one of the first systematic comparisons of major LLMs specifically in fetal CNS MRI interpretation, a domain that combines radiological expertise with prenatal counseling complexity. Text-based reasoning — using MRI report descriptions rather than raw images — was the evaluated modality, reflecting how LLMs might realistically be deployed to support clinicians who receive written radiology reports rather than directly reading images. Full performance data were not available in the published abstract, but the study's design — retrospective, single-center, with diagnostic accuracy as the primary endpoint — provides a methodologically grounded framework for evaluating LLM clinical utility beyond general medical knowledge benchmarks. The inclusion of DeepSeek alongside ChatGPT and Gemini reflects the model's growing presence in comparative medical AI evaluation. Results will contribute to the evidence base for whether general-purpose LLMs can support specialized prenatal diagnostic workflows. ---
Subscribe to read full wiki analysis
Full wiki content — including source provenance, contradiction tracking, related pages, and analyst notes — is available to Subscriber members.
View Plans & SubscribeNew users get a 7-day trial — no credit card required.