TRACE: Deployable Tree-Relational Structure Enhancement for Oncology LLMs

2026-09-20

AI Summary

This study addresses the lack of explicit medical structural support in large foundation models for oncology prediction by proposing a deployable tree-structured relation enhancement framework. Methodologically, it decouples offline structure learning from online lightweight inference, optimizing the tree topology via language modeling loss and employing a zero-shot unsupervised task-adaptive evidence retrieval mechanism to achieve compact evidence selection aligned with clinical reasoning. Experimental results demonstrate that the proposed approach significantly outperforms general RAG and GraphRAG baselines across multiple oncology tasks, effectively improving predictive accuracy while supporting dynamic structural updates and result auditability.

Abstract

Large language models are increasingly used in oncology applications, but their predictions are often weakly grounded in explicit medical structure. We present TRACE, a deployable tree-relational enhancement framework for oncology LLMs. TRACE separates expensive offline structure learning from lightweight online inference: oncology concepts and relations are organized into an updatable tree-relational structure, refined using LM-loss-derived evidence, and retrieved at inference time as compact prompt evidence. This design supports task-adaptive evidence selection without requiring supervised labels in the zero-shot setting. Across ten oncology classification tasks and one MedQuAD CancerGov QA benchmark, TRACE improves both label-free evaluation and supervised fine-tuning. Additional analyses show that TRACE improves over vanilla RAG and generic GraphRAG, remains useful under leakage-controlled METABRIC inputs, and produces interpretable evidence paths aligned with clinical reasoning. These results suggest that explicit, updatable medical structure is a practical path toward more accurate and auditable oncology LLM deployment.

Problem

• Large Language Models
• Oncology
• Medical Knowledge Structure
• Grounded Predictions
• Clinical Reasoning

Innovation

• Tree-Relational Structure
• Oncology LLMs
• Zero-Shot Evidence Selection
• GraphRAG
• Interpretable Reasoning