3DMedAgent: Unified Perception-to-Understanding for 3D Medical Analysis

TL;DR

3DMedAgent bridges the gap between 2D multimodal language models and 3D CT analysis by introducing a memory-augmented agent that moves from global perception to structured understanding. It decomposes 3D tasks via Organ-Aware Memory Initialization (OAMI), Coarse-to-Fine Lesion Targeting (CFLT), and Think-with-1-Slice Loop (T1S-Loop), accumulating compact evidence in a long-term memory to support multi-step reasoning. The DeepChestVQA benchmark is proposed to evaluate unified perception-to-understanding across thoracic CT, and experiments demonstrate that 3DMedAgent outperforms general, medical, and 3D-specific MLLMs across 40+ tasks with robust generalization. This approach offers a scalable pathway toward general-purpose 3D clinical assistants, while acknowledging limitations in purely zero-shot policy optimization and the need for strict clinical validation and supervision.

Abstract

3D CT analysis spans a continuum from low-level perception to high-level clinical understanding. Existing 3D-oriented analysis methods adopt either isolated task-specific modeling or task-agnostic end-to-end paradigms to produce one-hop outputs, impeding the systematic accumulation of perceptual evidence for downstream reasoning. In parallel, recent multimodal large language models (MLLMs) exhibit improved visual perception and can integrate visual and textual information effectively, yet their predominantly 2D-oriented designs fundamentally limit their ability to perceive and analyze volumetric medical data. To bridge this gap, we propose 3DMedAgent, a unified agent that enables 2D MLLMs to perform general 3D CT analysis without 3D-specific fine-tuning. 3DMedAgent coordinates heterogeneous visual and textual tools through a flexible MLLM agent, progressively decomposing complex 3D analysis into tractable subtasks that transition from global to regional views, from 3D volumes to informative 2D slices, and from visual evidence to structured textual representations. Central to this design, 3DMedAgent maintains a long-term structured memory that aggregates intermediate tool outputs and supports query-adaptive, evidence-driven multi-step reasoning. We further introduce the DeepChestVQA benchmark for evaluating unified perception-to-understanding capabilities in 3D thoracic imaging. Experiments across over 40 tasks demonstrate that 3DMedAgent consistently outperforms general, medical, and 3D-specific MLLMs, highlighting a scalable path toward general-purpose 3D clinical assistants.Code and data are available at \href{https://github.com/jinlab-imvr/3DMedAgent}{https://github.com/jinlab-imvr/3DMedAgent}.

Figures (7)

• Figure 1: The illustration of existing tasks and methods in 3D medical image analysis. 3DMedAgent adaptively invokes visual tools for task-specific perception and analysis, and summarizes the outputs as compact evidence stored in shared memory for subsequent reasoning.
• Figure 2: The overall framework of 3DMedAgent. A 2D MLLM agent iteratively interacts with heterogeneous tools, distills their outputs into structured evidence, and updates a shared memory for reliable 3D medical image analysis. Subtitle colors follow the same agentic action-type scheme, and colors denote agentic action types: tool use (purple), memory (blue), and reasoning (red).
• Figure 3: 3DMedAgent's reasoning trajectories across representative cases. 3DMedAgent addresses 3D CT queries by progressively gathering and reusing evidence in shared memory, illustrated on (a) measurement, (b) recognition, and (c) medical reasoning tasks.
• Figure 4: Performance comparison across source datasets in DeepTumorVQA.
• Figure 5: Performance comparison across different organs.