The Path to Self-Evolving Clinical Systems: Scaling Medical Agents from Assistance to Autonomy

2026-07-13

AI Summary

This work addresses the limitations of current medical AI agents in task generalization, environmental interaction, and continuous evolution by proposing a self-evolving autonomous agent framework designed for real-world clinical deployment. The framework models the agent as a partially observable sequential decision-making system and introduces a novel three-tier autonomy taxonomy. Centered on clinical environment expansion, it integrates large language models and vision-language models with an agent training sandbox, tool integration, clinical data interfaces (e.g., PACS, EHR, FHIR), and test-time compute techniques to enable progression from assistive to fully autonomous operation. Through a systematic review of over 300 studies, the paper delineates multimodal clinical use cases and identifies critical challenges—including hallucination, cascading failures, and fairness—providing a roadmap for building trustworthy, scalable, self-evolving medical imaging systems.

Abstract

The growing ability of large language models and vision language models to jointly interpret and reason over images and text is reshaping medical agents, moving them from task specific predictors toward autonomous systems that perceive, reason, plan, remember, and act in clinical environments. This work departs from the capability first perspective of existing literature and instead begins from clinical deployment, asking what tasks, contamination resistant benchmarks, and interactive training environments are required before medical agents can be trusted in practice. Medical agents are formalized as sequential decision making systems under partial observability, together with a three level autonomy taxonomy spanning assisted, cooperative, and fully autonomous operation. The field is organized along a unified scaling spine consisting of framework scaling, capability scaling, and environment scaling. Within this framework, clinical environment scaling, the integration of tools, data, and clinical gyms, is identified as the most actionable yet underexplored direction for agents operating in PACS, EHR, and FHIR ecosystems. Clinical self evolution, where agents improve through interaction with their environments rather than parameter scaling alone, is further positioned as a key research frontier, drawing insights from self improving agents, agent gyms, and test time compute scaling. Applications across radiology, pathology, ophthalmology, and hospital workflows are examined together with deployment challenges including hallucination, cascading failures, and fairness. By consolidating more than 300 references, with particular emphasis on advances from 2025 to 2026, this work provides a roadmap toward trustworthy, self improving medical imaging systems for real clinical practice.

Problem

• medical agents
• clinical autonomy
• self-evolving systems
• clinical deployment
• trustworthy AI

Innovation

• clinical self-evolution
• medical agents
• autonomy taxonomy
• environment scaling
• sequential decision making