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

arXiv: https://arxiv.org/abs/2604.16914 · HTML: https://mlsanity.com/paper/2604.16914/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 10 of its 17 references that are in the map.
- First posted on arXiv 18 Apr 2026.

## Builds on (most cited first, 10 of 10)

- [Segment Anything](https://mlsanity.com/paper/2304.02643/) (2023)
- [Toolformer: Language Models Can Teach Themselves to Use Tools](https://mlsanity.com/paper/2302.04761/) (2023)
- [LLaVA-Med: Training a Large Language-and-Vision Assistant for Biomedicine in One Day](https://mlsanity.com/paper/2306.00890/) (2023)
- [Which Tasks Should Be Learned Together in Multi-task Learning?](https://mlsanity.com/paper/1905.07553/) (2019)
- [Multi-Task Learning with Deep Neural Networks: A Survey](https://mlsanity.com/paper/2009.09796/) (2020)
- [Segment Anything in Medical Images](https://mlsanity.com/paper/2304.12306/) (2023)
- [On the Challenges and Perspectives of Foundation Models for Medical Image Analysis](https://mlsanity.com/paper/2306.05705/) (2023)
- [USFM: A Universal Ultrasound Foundation Model Generalized to Tasks and Organs towards Label Efficient Image Analysis](https://mlsanity.com/paper/2401.00153/) (2023)
- [TinyUSFM: Towards Compact and Efficient Ultrasound Foundation Models](https://mlsanity.com/paper/2510.19239/) (2025)
- [Baseline Method of the Foundation Model Challenge for Ultrasound Image Analysis](https://mlsanity.com/paper/2602.01055/) (2026)

## Abstract (arXiv)

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 FMCUIA 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.