Expert-level localization of fine-grained cardiac structures across the transthoracic echocardiogram
Artificial intelligence models for echocardiography interpret studies without identifying the structures that support their outputs. We show that this spatial grounding does not emerge from the pretraining of current vision-language models. General-purpose and medical vision-language models and an open-vocabulary detector recalled 3% or fewer of cardiac structures and scored below a box that never sees the image. EchoLens, a detector trained on 5,912 expert-annotated structures and 4,219 Doppler signals from 136 patients, recalled 87% and localized structures as reliably as a blinded expert reader (box centre within the reference in 95.7% and 95.3% of cases). It covers 21 structures across 11 transthoracic views and 11 spectral Doppler signal types. In external data, two physicians judged 81-92% of its detections correct across all 11 views, and it identified the Doppler signal type in 83% of frames. EchoLens provides the spatial layer that grounded echocardiographic interpretation requires.
Authors
- Hesham Sheashaa (ORCID: https://orcid.org/0009-0000-4740-5388)
- Bradley James Erickson (ORCID: https://orcid.org/0000-0001-7926-6095)
- Elham Mahmoudi (ORCID: https://orcid.org/0000-0001-5160-0506)
- Chadi Ayoub (ORCID: https://orcid.org/0000-0001-5877-7478)
- Nicolò Pecco (ORCID: https://orcid.org/0000-0002-6367-7903)
- Imon Banerjee (ORCID: https://orcid.org/0000-0002-3327-8004)
- Chieh‐Ju Chao (ORCID: https://orcid.org/0000-0001-6155-0266)
- Reza Arsanjani (ORCID: https://orcid.org/0000-0001-7081-4286)
- Isabel G. Scalia (ORCID: https://orcid.org/0000-0002-6459-9767)
- Kamal Awad (ORCID: https://orcid.org/0000-0001-6707-1629)
- Martha A. Grogan (ORCID: https://orcid.org/0000-0002-2541-4440)
- Juan Farina (ORCID: https://orcid.org/0000-0002-5824-8485)
- Fatmaelzahraa E. Abdelfattah (ORCID: https://orcid.org/0009-0008-0758-1086)
- Nima Baba Ali (ORCID: https://orcid.org/0009-0004-3979-0609)
- Sogol Attaripour Esfahani (ORCID: https://orcid.org/0000-0003-3777-7369)
- Garvan Kane
- Jae K. Oh
Institutions
- Cleveland Clinic (US)
- Mayo Clinic (US)
- Vita-Salute San Raffaele University (IT)
- Mayo Clinic Hospital (US)
Publication Details
- Journal
- medRxiv
- Published
- 2026-11-12
- DOI
- https://doi.org/10.64898/2026.11.10.26367643
- Primary Topic
- Artificial Intelligence in Healthcare and Education
- Type
- preprint