PRECISE-AS: Personalized Reinforcement Learning for Efficient Point-of-Care Echocardiography in Aortic Stenosis Diagnosis

2025-09-02

AI Summary

Transthoracic echocardiography (TTE) for aortic stenosis (AS) diagnosis faces limited accessibility and high operator dependency in rural and resource-constrained settings. Method: We propose a reinforcement learning–based active video acquisition framework that abandons fixed-view protocols; instead, it dynamically models information gain across clinical video sequences to select, in real time, the most discriminative imaging views—enabling personalized, on-demand point-of-care ultrasound acquisition. Contribution/Results: Our approach tightly couples active feature acquisition with the clinical AS classification objective, yielding an end-to-end decision policy. Evaluated on 2,572 real-world patient studies, it achieves 80.6% AS classification accuracy using only 47% of conventional video frames. This substantially reduces reliance on high-end equipment and operator expertise, thereby enhancing diagnostic efficiency and resource utilization in primary care settings.

Technology Category

• Machine Learning: Active Learning
• Constraint Satisfaction and Optimization: Constraint Learning and Acquisition
• Computer Vision: Video Understanding & Activity Analysis

Application Category

• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Agentic search

Abstract

Aortic stenosis (AS) is a life-threatening condition caused by a narrowing of the aortic valve, leading to impaired blood flow. Despite its high prevalence, access to echocardiography (echo), the gold-standard diagnostic tool, is often limited due to resource constraints, particularly in rural and underserved areas. Point-of-care ultrasound (POCUS) offers a more accessible alternative but is restricted by operator expertise and the challenge of selecting the most relevant imaging views. To address this, we propose a reinforcement learning (RL)-driven active video acquisition framework that dynamically selects each patient's most informative echo videos. Unlike traditional methods that rely on a fixed set of videos, our approach continuously evaluates whether additional imaging is needed, optimizing both accuracy and efficiency. Tested on data from 2,572 patients, our method achieves 80.6% classification accuracy while using only 47% of the echo videos compared to a full acquisition. These results demonstrate the potential of active feature acquisition to enhance AS diagnosis, making echocardiographic assessments more efficient, scalable, and personalized. Our source code is available at: https://github.com/Armin-Saadat/PRECISE-AS.

Problem

• Improving aortic stenosis diagnosis efficiency
• Reducing required echocardiography video quantity
• Personalizing point-of-care ultrasound acquisition strategy

Innovation

• Reinforcement learning-driven active video acquisition
• Dynamically selects most informative echocardiography videos
• Optimizes accuracy and efficiency with personalized approach