AI-assisted echocardiography training

Supporting better acquisition through structured feedback

EchoLearn connects echocardiography practice with AI-assisted image assessment and visual feedback. It is being developed to support training and supervised practice—not to make diagnoses or replace clinical judgement.

Developed by the Intelligent Sensing and Vision Group at the THRIVE Centre, University of West London, EchoLearn builds on research from the UNITY programme, supported by the British Heart Foundation.

Who it supportsLearners, sonographers, educators, and clinical teams
What it focuses onImage acquisition, feedback, and supervised training
Where it standsResearch prototype in supervised evaluation

Why EchoLearn

Making feedback part of the learning process

A skill that takes practice

Obtaining useful echocardiographic views depends on probe position, orientation, and adapting to each patient's anatomy. Learners build these skills through practice and expert supervision.

Support that is structured and timely

Service pressures can make consistent feedback difficult. EchoLearn explores whether immediate, objective visual feedback can help learners and supervisors recognise image quality and discuss possible adjustments.

From scan to review

How EchoLearn works

The platform is designed to support a practical learning cycle, with educators and clinicians remaining central to supervision.

Acquire

The learner practises obtaining echocardiographic views using an ultrasound system.

Assess

AI tools analyse the video and generate image-quality or geometric indicators.

Guide

Visual feedback supports discussion and acquisition decisions with a supervisor.

Review

Session information can support reflection and research into the training workflow.

Platform and progress

A research platform in development

EchoLearn combines AI-based image assessment with echocardiography acquisition and operator feedback. Capabilities are being developed and evaluated in stages.

Research prototype

Available in the prototype

  • Live echocardiographic video input
  • AI-based image assessment
  • On-screen visual feedback
  • Session capture and analysis
Under evaluation

Being explored with learners

  • Supervised use in scanning practice
  • Usability and fit with training
  • How operators respond to feedback
  • Potential support for novice image acquisition
In development

Future directions

  • Longitudinal trainee progress tracking
  • Deployment across ultrasound systems

Clinical evaluation

Learning from supervised practice

A Phase 1 evaluation began in August 2026 in an echocardiography training setting at Hammersmith Hospital. Medical students take part in supervised scanning sessions. The evaluation is exploring usability, how learners respond to feedback, and whether real-time guidance helps novice operators acquire better images.

A later controlled evaluation is intended to assess acquisition quality, acquisition time, reproducibility, and downstream measurements. Results have not yet been reported, and improved clinical performance has not yet been demonstrated.

Current focusFeasibility, usability, and learner response in supervised training.

Part of a wider research ecosystem

Connecting research, data, and training

EchoLearn sits within THRIVE and UNITY's wider echocardiography research environment. Related resources include:

EchoForge

Reusable AI tools and research resources for echocardiography.

UNITY Biobank

Echocardiography data and annotation infrastructure supporting research.

People and organisations

Partnerships

Research lead

THRIVE Centre, University of West London

Clinical partner

National Heart and Lung Institute

Evaluation setting

Hammersmith Hospital

Funding

British Heart Foundation Programme Grant RG/F/22/110059

Work with us

Collaboration and contact

We welcome conversations with echocardiography educators, NHS training programmes, and researchers interested in image quality, operator guidance, and independent evaluation.

Contact Massoud Zolgharni or [email protected] to discuss collaboration.

EchoLearn is a research platform. It does not provide diagnostic interpretation or automated clinical decision-making. Any use in clinical care requires appropriate evaluation and validation.