Unified Ultrasound Intelligence Toward an End-to-End Agentic System

TL;DR

The paper introduces USTri, a three-stage ultrasound analysis pipeline that generalizes across organs, tasks, and devices, enabling end-to-end clinical workflow automation with high accuracy.

Contribution

It proposes a novel tri-stage framework combining a universal generalist, dataset-specific fine-tuning, and workflow-mimicking inference, advancing ultrasound AI capabilities.

Findings

Achieves state-of-the-art performance on FMC_UIA dataset across multiple tasks.

Produces clinically structured, interpretable reports with high accuracy.

Demonstrates robustness to device and protocol variability.

Abstract

Clinical ultrasound analysis demands models that generalize across heterogeneous organs, views, and devices, while supporting interpretable workflow-level analysis. Existing methods often rely on task-wise adaptation, and joint learning may be unstable due to cross-task interference, making it hard to deliver workflow-level outputs in practice. To address these challenges, we present USTri, a tri-stage ultrasound intelligence pipeline for unified multi-organ, multi-task analysis. Stage I trains a universal generalist USGen on different domains to learn broad, transferable priors that are robust to device and protocol variability. To better handle domain shifts and reach task-aligned performance while preserving ultrasound shared knowledge, Stage II builds USpec by keeping USGen frozen and finetuning dataset-specific heads. Stage III introduces USAgent, which mimics clinician workflows by orchestrating USpec specialists for multi-step inference and deterministic structured reports. On the FMC\_UIA validation set, our model achieves the best overall performance across 4 task types and 27 datasets, outperforming state-of-the-art methods. Moreover, qualitative results show that USAgent produces clinically structured reports with high accuracy and interpretability. Our study suggests a scalable path to ultrasound intelligence that generalizes across heterogeneous ultrasound tasks and supports consistent end-to-end clinical workflows. The code is publicly available at: https://github.com/MacDunno/USTri.