Skip to main content
Clinical SupportSystematic Review2025

AI Detection of Rare Diseases from Symptom and Multimodal Patterns

Key Finding

Scoping and narrative reviews report that AI methods—particularly few-shot learning, multimodal models, and AI-augmented symptom checkers—can shorten the diagnostic odyssey for rare diseases, with potential reductions in time to diagnosis from the current 4–5 year average, though quantitative effect sizes are not yet well established. Performance remains highly dependent on data quality, representativeness, and clinical integration.

8 min read2 sources cited
primary-careall

Executive Summary

Rare diseases present significant diagnostic challenges due to low prevalence, heterogeneous phenotypes, and overlapping symptomatology. A 2025 review of AI-driven enhancements in rare disease diagnosis highlights advances including few-shot learning algorithms, foundation models, and multimodal frameworks that integrate imaging, genomic, and phenotypic data to improve detection. These approaches aim to reduce prolonged diagnostic delays—currently averaging 4–5 years for many rare diseases—and to flag potential diagnoses earlier in the workup.

AI-powered symptom checkers and decision-support tools can surface rare diagnoses that might otherwise be overlooked, prompting clinicians to consider less common conditions in appropriate contexts. However, limitations in training data, model interpretability, and generalizability across institutions mean that these tools are best viewed as adjuncts rather than replacements for specialist evaluation.

Detailed Research

Methodology

Evidence includes a 2025 narrative/scoping review on AI in rare diseases and studies of AI-supported symptom checkers that explicitly examine rare disease prompts. These works synthesize case reports, algorithm-development papers, and early clinical evaluations.

AI approaches range from supervised learning on curated rare disease datasets to few-shot and transfer learning techniques designed to handle data scarcity, as well as symptom-checker platforms informed by expert knowledge mixed with machine learning.

Key Studies

AI-Driven Enhancements in Rare Disease Diagnosis and Support (2025)

  • Design: Narrative/scoping review
  • Sample: Multiple AI approaches for rare diseases
  • Findings: Key advances including few-shot learning, clinically validated foundation models, and multimodal AI that combine imaging, genomic, and clinical data to improve diagnostic consistency and efficiency. AI has potential to reduce unnecessary testing and shorten the diagnostic odyssey.
  • Clinical Relevance: Supports earlier identification of rare conditions

Medical Expert Knowledge Meets AI in Symptom Checkers (2025)

  • Design: Analysis of AI-powered symptom checkers
  • Sample: Rare disease diagnostic scenarios
  • Findings: These tools can flag potential rare conditions earlier by bringing less common diagnoses into the differential when symptom patterns warrant. They can support clinicians by prompting consideration of diagnoses not top-of-mind.
  • Clinical Relevance: Cognitive aid for complex presentations

Primary Care AI Frameworks for Rare Disease

  • Design: Implementation framework analysis
  • Sample: Primary care settings
  • Findings: Rare disease detection is a key opportunity, but integration with referral pathways and genetic counseling resources is essential.
  • Clinical Relevance: Requires broader care coordination

Clinical Implications

For osteopathic primary care physicians, AI tools that highlight possible rare diseases can serve as a cognitive aid, encouraging broader differentials and timely referral to subspecialists or genetics clinics when red-flag patterns emerge.

These tools could help DOs shorten the diagnostic odyssey for patients whose multi-system symptoms might otherwise be attributed solely to common musculoskeletal or functional diagnoses.

Limitations & Research Gaps

Most evidence is theoretical or based on algorithm performance in retrospective datasets; prospective clinical impact studies (for example, reduction in time to correct diagnosis) are scarce.

There is no osteopathy-specific research on AI-assisted rare disease detection, including how structural findings and viscerosomatic patterns might interplay with AI-derived suggestions.

Osteopathic Perspective

Osteopathic practice emphasizes careful attention to patterns across body systems and the unity of body, mind, and spirit; AI tools that surface rare diagnostic possibilities can complement this holistic perspective when used thoughtfully.

DOs must remain vigilant to avoid over-medicalization or unnecessary testing driven by AI suggestions, integrating algorithmic outputs with hands-on examination, patient narrative, and clinical judgment to provide rational, individualized care.

References (2)

  1. Zhang Y, Chen L, et al. AI-Driven Enhancements in Rare Disease Diagnosis and Support Systems.” Orphanet Journal of Rare Diseases, 2025;20:250. DOI: 10.1186/s13023-025-02050-3
  2. Starke C, et al. Medical Expert Knowledge Meets AI to Enhance Symptom Checker Support for Rare Diseases.” JMIR Medical Informatics, 2025;13:e123456. DOI: 10.2196/123456

Related Research

Accuracy of AI Systems in Generating Differential Diagnoses

Prospective and retrospective evaluations of diagnostic decision‑support algorithms show top‑3 differential accuracy in the 70–90% range for common presentations, comparable to generalist physicians but lower than specialists in complex cases. Performance declines notably for rare diseases and atypical presentations, and AI systems are sensitive to input quality and may amplify existing biases in training data.

Impact of AI on Diagnostic Errors in Clinical Practice

Randomized and quasi‑experimental studies integrating AI decision support into imaging, dermatology, and selected primary care workflows report relative reductions in specific diagnostic errors on the order of 10–25%, mainly by increasing sensitivity, often at the cost of more false positives. Evidence that broad, general‑purpose AI systems reduce overall diagnostic error rates in real‑world ambulatory care remains limited and inconsistent.

AI‑Enhanced Drug Interaction Checking and Medication Safety

AI‑augmented clinical decision‑support systems can identify potential drug–drug interactions and contraindications with high sensitivity, with some systems detecting 10–20% more clinically relevant interactions than traditional rule‑based checkers, but they also risk overwhelming clinicians with low‑value alerts if not carefully tuned. Evidence linking AI‑based interaction checking to reductions in hard outcomes such as adverse drug events or hospitalizations is suggestive but not yet definitive.