Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

EchoAgent: Guideline-Centric Reasoning Agent for Echocardiography Measurement and Interpretation

Matin Daghyani, Lyuyang Wang, Nima Hashemi, Bassant Medhat, Baraa Abdelsamad, Eros Rojas Velez, XiaoXiao Li, Michael Y. C. Tsang, Christina Luong, Teresa S. M. Tsang, Purang Abolmaesumi

TL;DR

EchoAgent addresses the need for video-level, guideline-based reasoning in echocardiography by orchestrating domain-specific tools under an LLM controller. It introduces a measurement-feasibility predictor and guideline retrieval to ground outputs in clinically validated standards, enabling interpretable, end-to-end analysis of echo videos. On a curated benchmark, EchoAgent achieves accurate, auditable results while exposing intermediate reasoning and visual evidence, highlighting the approach's potential to improve trust and usability in bedside ultrasound. The work demonstrates a practical path toward trustworthy AI in cardiac ultrasound through structured tool use and full video-level automation.

Abstract

Purpose: Echocardiographic interpretation requires video-level reasoning and guideline-based measurement analysis, which current deep learning models for cardiac ultrasound do not support. We present EchoAgent, a framework that enables structured, interpretable automation for this domain. Methods: EchoAgent orchestrates specialized vision tools under Large Language Model (LLM) control to perform temporal localization, spatial measurement, and clinical interpretation. A key contribution is a measurement-feasibility prediction model that determines whether anatomical structures are reliably measurable in each frame, enabling autonomous tool selection. We curated a benchmark of diverse, clinically validated video-query pairs for evaluation. Results: EchoAgent achieves accurate, interpretable results despite added complexity of spatiotemporal video analysis. Outputs are grounded in visual evidence and clinical guidelines, supporting transparency and traceability. Conclusion: This work demonstrates the feasibility of agentic, guideline-aligned reasoning for echocardiographic video analysis, enabled by task-specific tools and full video-level automation. EchoAgent sets a new direction for trustworthy AI in cardiac ultrasound.

EchoAgent: Guideline-Centric Reasoning Agent for Echocardiography Measurement and Interpretation

TL;DR

EchoAgent addresses the need for video-level, guideline-based reasoning in echocardiography by orchestrating domain-specific tools under an LLM controller. It introduces a measurement-feasibility predictor and guideline retrieval to ground outputs in clinically validated standards, enabling interpretable, end-to-end analysis of echo videos. On a curated benchmark, EchoAgent achieves accurate, auditable results while exposing intermediate reasoning and visual evidence, highlighting the approach's potential to improve trust and usability in bedside ultrasound. The work demonstrates a practical path toward trustworthy AI in cardiac ultrasound through structured tool use and full video-level automation.

Abstract

Purpose: Echocardiographic interpretation requires video-level reasoning and guideline-based measurement analysis, which current deep learning models for cardiac ultrasound do not support. We present EchoAgent, a framework that enables structured, interpretable automation for this domain. Methods: EchoAgent orchestrates specialized vision tools under Large Language Model (LLM) control to perform temporal localization, spatial measurement, and clinical interpretation. A key contribution is a measurement-feasibility prediction model that determines whether anatomical structures are reliably measurable in each frame, enabling autonomous tool selection. We curated a benchmark of diverse, clinically validated video-query pairs for evaluation. Results: EchoAgent achieves accurate, interpretable results despite added complexity of spatiotemporal video analysis. Outputs are grounded in visual evidence and clinical guidelines, supporting transparency and traceability. Conclusion: This work demonstrates the feasibility of agentic, guideline-aligned reasoning for echocardiographic video analysis, enabled by task-specific tools and full video-level automation. EchoAgent sets a new direction for trustworthy AI in cardiac ultrasound.

Paper Structure

This paper contains 16 sections, 2 figures, 3 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Related Work
  3. Foundation Models for Echocardiography
  4. Agentic AI in Medicine
  5. EchoAgent
  6. Overview
  7. Agentic Design
  8. Tools Integration
  9. Evaluation
  10. Dataset
  11. Implementation Details
  12. EchoAgent Evaluation
  13. Tools Evaluation
  14. Ablation Study
  15. Discussion
  16. ...and 1 more sections

Figures (2)

  • Figure 1: Overview of the EchoAgent framework. EchoAgent is a guideline-centric reasoning system that integrates specialized visual tools under the orchestration of an LLM. Given a clinician’s natural-language query and an echo video, the LLM orchestrator dynamically invokes tools for phase detection, feasibility prediction, measurement segmentation, and guideline retrieval to produce interpretable, guideline-aligned answers. The system iteratively reasons over intermediate observations to ensure that all measurements are visually grounded and clinically validated.
  • Figure 2: EchoAgent Interaction Flow. Examples with visualizations of how EchoAgent handles a user query through reasoning and interaction with different tools. The example on the left is an easier case; the example on the right requires derived calculations.