AI

RadOnc-Agent: An LLM-Orchestrated Framework for AI Workflows Across the Radiotherapy Care Pathway

Read full article at source ↗

Based on a feed excerpt; full article content has not been acquired. Source evidence retrieved 08/10/2026, 02:18:24 UTC.

Briefings

RadOnc-Agent is an LLM-orchestrated framework that formalizes radiotherapy into four clinical phases using 26 callable functions via a conversational interface. Evaluated on 2,600 single-function requests and 120 workflow instances from 60 de-identified patients, the system selected the intended function in 98.79% of single-executions and completed 96.67% of real-patient workflows. Key ablation studies showed that removing longitudinal state reduced cross-stage completion to 84.00%, while disabling schema and identity validation increased mismatched backend dispatch to 95.28%.

Key points

• RadOnc-Agent uses 26 callable functions to formalize radiotherapy into four clinical phases via a conversational interface.
• System accuracy metrics: 98.79% selection rate for single functions and 96.67% completion rate for real-patient workflows.

Context

This research addresses fragmentation in radiotherapy workflows by proposing an agentic AI framework that unifies decision-making, planning, and adaptation across longitudinal patient care paths.

Limitations

Findings establish technical feasibility of LLM orchestration but do not confirm clinical correctness, utility, or prospective benefit. Data includes 60 de-identified patient records and synthetic scenarios; specific city or region location is not supported by the evidence.

Generated by Bittrees-hosted qwen35-2b sha256:aaf42c8b7c3c on 26/10/2026, 03:54:48 UTC. Based on publisher-provided feed evidence; may not cover the full original article. Check the source for details.

View archived briefing on IPFS
bafybeiaznqto4r2kwlskbthaa4tuxwtqycg57hdvgoe3slndufblgeo6wq