LiteOdyssey: A Lightweight Reasoning AI Agent for Interpretable Rare-Disease Diagnosis

2026-06-14

AI Summary

Existing medical AI systems for rare disease diagnosis often rely on large-scale datasets, multi-agent architectures, or extensive retrieval databases, resulting in deployment challenges, poor auditability, and high maintenance costs. This work proposes a lightweight, interpretable single-agent diagnostic framework that operates without fine-tuning, multi-agent coordination, or large case repositories. By integrating clinical genetics workflows, Policy Iteration with Human Feedback (PIHF), dynamic invocation of public biomedical tools, and a reasoning-guided language model mechanism, the system achieves significantly enhanced transparency and deployability. On the LIRICAL and PhenoPacket Store benchmarks, the method attains a Recall@1 of 59.3%, rising to 60.7% on a more challenging PhenoPacket subset—substantially outperforming a no-tool baseline model, which achieves only 10.7%.

Abstract

Most medical AI systems improve by scaling additional machinery: more fine-tuning data, more agents, and/or larger retrieval databases. In rare-disease diagnosis, however, such scaling can produce systems that are difficult to deploy, audit, and maintain. We asked whether state-of-the-art diagnostic performance could instead be achieved by extending the reasoning chain of a single AI agent: guiding it with a diagnostic policy, developed through human-AI collaboration and augmenting with freely available biomedical tools. We introduce LiteOdyssey, a lightweight rare-disease diagnostic framework that guides reasoning language model through a clinical genetics workflow. This framework was developed through Policy Iteration with Human Feedback (PIHF) and uses dynamic access to public biomedical tools. On two challenging benchmarks that provide only patient clinical features, LiteOdyssey achieved state-of-the-art performance, with an overall disease Recall@1 of 59.3% over the combined 1,243 cases of LIRICAL (n = 370) and the PhenoPacket Store (n = 873). Both benchmarks have a high proportion of ultra-rare disease (a prevalence below 1 in 1,000,000, with ultra-rare shares of approximately 45% and 52.8%, respectively). On the more difficult PhenoPacket subset, where causal diseases were not mapped to Orphanet in our rarity-mapping pipeline, LiteOdyssey achieved 60.7% Recall@1, compared with 10.7% for the same baseline model (GPT-5.4) without tools. This performance was achieved without fine-tuning, multi-agent ensembles, or a large case-retrieval database. Gains were also observed in the following: on cases never seen during development, on a private cohort of real-world rare disease patients, and on a smaller open-weights model. LiteOdyssey suggests a path toward rare-disease AI systems that are accurate, easier to deploy, and more transparent for physician review.

Problem

• rare-disease diagnosis
• lightweight AI
• interpretable AI
• clinical deployment
• diagnostic reasoning

Innovation

• lightweight AI agent
• reasoning chain extension
• Policy Iteration with Human Feedback (PIHF)
• public biomedical tool integration
• interpretable rare-disease diagnosis