---
title: "EchoQuality: Automated echocardiography image quality assessment and its effect on AI performance"
id: "medrxiv-12-automated-echocardiography-image-quality-assessment-and-its-association-with-ai"
canonical_url: "https://medichelpline.com/clinical-feed/medrxiv-12-automated-echocardiography-image-quality-assessment-and-its-association-with-ai"
content_type: "clinical_feed_article"
specialty: "Cardiology"
source_name: "medRxiv (Clinical Preprints)"
source_url: "https://www.medrxiv.org/content/10.64898/2026.11.19.26368489v1?rss=1"
published_at: "2026-11-20T12:00:00.000Z"
evidence_level: "Verified Feed"
license: "CC-BY-NC-4.0 / Informational Use"
---
# EchoQuality: Automated echocardiography image quality assessment and its effect on AI performance
## Provenance & Clinical Metadata
- **Canonical URL:** https://medichelpline.com/clinical-feed/medrxiv-12-automated-echocardiography-image-quality-assessment-and-its-association-with-ai
- **Specialty:** [Cardiology](https://medichelpline.com/clinical-feed/cardiology.md)
- **Primary Source:** medRxiv (Clinical Preprints)
- **Source URL:** [Original Journal Publication](https://www.medrxiv.org/content/10.64898/2026.11.19.26368489v1?rss=1)
- **Published At:** 2026-11-20T12:00:00.000Z
- **Evidence Rating:** Verified Feed
## Executive GIST (TL;DR)
- A multidisciplinary team developed **EchoQuality**, a deep-learning model trained to assign automated quality scores to echocardiography videos after creating an operational definition of image quality through co-creation. - Investigators annotated 6,474 videos for development and internal testing, and set aside 1,000 videos for external testing; the paper reports EchoQuality matched inter-observer agreement levels in internal comparisons. - In a prospective blinded best-clip selection experiment, the study team observed a 90.4% agreement between EchoQuality and an expert reviewer on clip selection. - The authors tested how image quality affected downstream **AI** tasks across 762,588 videos from 25,205 studies spanning five international cohorts; they evaluated 12 quantitative measurements, 5 moderate dysfunctions, and 6 cardiac disease detections. - High-quality studies produced superior AI performance in 97 of 122 evaluations (80%); median performance gaps between high- and low-quality videos were substantial across tasks, for example a 10% median R2 gap for measurements and 14% median sensitivity gap for dysfunction detection. - Performance penalties appeared consistent across different state-of-the-art models, datasets, and tasks, indicating image quality is a general determinant of automated analysis performance. - The authors developed a quality-aware aggregation strategy to mitigate low-quality inputs; this approach improved performance in 63 of 72 evaluations (88%) with a median relative gain of 5%. - EchoQuality has been made publicly available on GitHub; access to several raw datasets requires institutional approvals as described by the authors. - The paper reports ethics committee approval from Vall d'Hebron University Hospital and declares no competing interests; the study does not report certain granular algorithmic hyperparameters or model architectures in the abstract, which would require consulting the full text or code repository for details.
## Clinical Analysis & Structured Key Points
Automated echocardiography image quality assessment and its association with AI performance | medRxiv Skip to main content Automated echocardiography image quality assessment and its association with AI performance Pere Lopez-Gutierrez , Juan Garrido-Oliver , Mireia Bragulat-Arevalo , Juan Pablo Betancur-Rengifo , Nicholas Cauwenberghs , Alberto Morales-Galan , Marawan Elbatel , Machteld J. Boonstra , Francesca Ferrando , Horacio Majul , Lisandro Rivas Catoni , Santiago Martin Grieve , Stephan van der Zwaard , Yi Qin , Folkert W. Asselbergs , Tatiana Kuznetsova , Gisela Teixido-Tura , Alex P.W. Lee , Xiaomeng Li , Karim Lekadir , Laura Galian-Gay , Jose Rodriguez-Palomares , Ignacio Ferreira-Gonzalez , Lydia Dux-Santoy , View ORCID Profile Andrea Guala doi: https://doi.org/10.64898/2026.11.19.26368489 Pere Lopez-Gutierrez 1 Vall d Hebron Institut de Recerca, Barcelona, Spain; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Juan Garrido-Oliver 1 Vall d Hebron Institut de Recerca, Barcelona, Spain; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Mireia Bragulat-Arevalo 1 Vall d Hebron Institut de Recerca, Barcelona, Spain; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Juan Pablo Betancur-Rengifo 1 Vall d Hebron Institut de Recerca, Barcelona, Spain; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Nicholas Cauwenberghs 2 KU Leuven, Leuven, Belgium; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Alberto Morales-Galan 1 Vall d Hebron Institut de Recerca, Barcelona, Spain; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Marawan Elbatel 3 The Hong Kong University of Science and Technology, Hong Kong SAR, China; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Machteld J. Boonstra 4 Amsterdam University Medical Center, University of Amsterdam, Amsterdam, the Netherlands; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Francesca Ferrando 1 Vall d Hebron Institut de Recerca, Barcelona, Spain; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Horacio Majul 1 Vall d Hebron Institut de Recerca, Barcelona, Spain; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Lisandro Rivas Catoni 1 Vall d Hebron Institut de Recerca, Barcelona, Spain; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Santiago Martin Grieve 1 Vall d Hebron Institut de Recerca, Barcelona, Spain; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Stephan van der Zwaard 4 Amsterdam University Medical Center, University of Amsterdam, Amsterdam, the Netherlands; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Yi Qin 3 The Hong Kong University of Science and Technology, Hong Kong SAR, China; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Folkert W. Asselbergs 4 Amsterdam University Medical Center, University of Amsterdam, Amsterdam, the Netherlands; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Tatiana Kuznetsova 2 KU Leuven, Leuven, Belgium; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Gisela Teixido-Tura 1 Vall d Hebron Institut de Recerca, Barcelona, Spain; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Alex P.W. Lee 5 The Chinese University of Hong Kong (Department of Medicine & Therapeutics), Hong Kong, China; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Xiaomeng Li 3 The Hong Kong University of Science and Technology, Hong Kong SAR, China; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Karim Lekadir 6 Artificial Intelligence in Medicine Lab (BCN-AIM), Universitat de Barcelona, Spain Find this author on Google Scholar Find this author on PubMed Search for this author on this site Laura Galian-Gay 1 Vall d Hebron Institut de Recerca, Barcelona, Spain; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Jose Rodriguez-Palomares 1 Vall d Hebron Institut de Recerca, Barcelona, Spain; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Ignacio Ferreira-Gonzalez 1 Vall d Hebron Institut de Recerca, Barcelona, Spain; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Lydia Dux-Santoy 1 Vall d Hebron Institut de Recerca, Barcelona, Spain; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Andrea Guala 1 Vall d Hebron Institut de Recerca, Barcelona, Spain; Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Andrea Guala For correspondence: andrea.guala{at}vhir.org Abstract Info/History Metrics Data/Code Preview PDF Abstract Background Echocardiographic image quality influences clinical interpretation, yet its definition remains subjective and its impact on artificial intelligence (AI) systems has not been systematically evaluated. We developed EchoQuality, a deep-learning model for automated echocardiography quality assessment, and quantified the effect of image quality on AI performance across a broad range of tasks. Methods A multidisciplinary co-creation process established a definition of echocardiographic image quality, which was used to annotate 6,474 videos for EchoQuality development and internal testing, and 1,000 videos for external testing. The impact of image quality on the performance of state-of-the-art AI models was subsequently assessed for the prediction of 12 measurements, and the detection of 5 moderate dysfunctions and 6 cardiac diseases across five international cohorts totalling 762,588 videos from 25,205 echocardiography studies. In addition, a quality-aware aggregation strategy was developed to mitigate the impact of low-quality images. Results EchoQuality achieved a performance comparable to the inter-observer agreement, while the agreement with an expert in a prospective, blinded best-clip selection experiment was 90.4%. In downstream AI uses, high-quality studies resulted in the highest performance in 97 of 122 (80%) of evaluations. The median performance gaps between high- and low-quality videos were 10% [5-20%] in the prediction of quantitative measurements (R2), 14% [7-20%] in the identification of moderate dysfunctions (sensitivity), 12% [3-20%] in the detection of diseases (sensitivity), 12% [9-15%] in view identification (accuracy), and 3% [2-5%] in left-ventricular segmentation (Dice score). Performance gaps were consistent across state-of-the-art models, datasets, and tasks. Finally, a quality-aware aggregation strategy improved performance in 88% (63 of 72) of the evaluations, with a median relative improvement of 5% [1-15%]. Conclusions EchoQuality provides an automated assessment of echocardiographic image quality, a major determinant of AI performance. Incorporating quality-awareness may improve the robustness and clinical reliability of echocardiographic AI systems. Competing Interest Statement The authors have declared no competing interest. Author Declarations I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained. Yes The details of the IRB/oversight body that provided approval or exemption for the research described are given below: Ethics committee of the Vall d'Hebron University Hospital gave ethical approval for this work. I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals. Yes I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance). Yes I have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable. Yes Data Availability Access to HUVH, PWH, AUMC and FLEMENGHO raw datasets will be granted on approval by the institutional review boards and data sharing committees. Details on how to request access are available from the corresponding author on reasonable request. MIMIC-IV-Echo and CAMUS are available at their respective web pages. EchoQuality is publicly available on our GitHub page https://github.com/CardiovascularImagingVallHebron/EchoQuality. https://github.com/CardiovascularImagingVallHebron/EchoQuality Funder Information Declared Instituto de Salud Carlos III , FORT23/00034 , CP24/00121 European Commission , DVPS, Grant agreement ID: 101213369 Biomedical Research Networking Center on Cardiovascular Diseases (CIBERCV) Generalitat de Catalunya , AGAUR-FI 2023 FI-1 00322 Joan Oró Copyright The copyright holder for this preprint is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. It is made available under a CC-BY-NC 4.0 International license . Back to top Previous Next Posted November 20, 2026. Download PDF Data/Code Email Thank you for your interest in spreading the word about medRxiv. NOTE: Your email address is requested solely to identify you as the sender of this article. Your Email * Your Name * Send To * Enter multiple addresses on separate lines or separate them with commas. You are going to email the following Automated echocardiography image quality assessment and its association with AI performance Message Subject (Your Name) has forwarded a page to you from medRxiv Message Body (Your Name) thought you would like to see this page from the medRxiv website. Your Personal Message CAPTCHA This question is for testing whether or not you are a human visitor and to prevent automated spam submissions. Share Automated echocardiography image quality assessment and its association with AI performance Pere Lopez-Gutierrez , Juan Garrido-Oliver , Mireia Bragulat-Arevalo , Juan Pablo Betancur-Rengifo , Nicholas Cauwenberghs , Alberto Morales-Galan , Marawan Elbatel , Machteld J. Boonstra , Francesca Ferrando , Horacio Majul , Lisandro Rivas Catoni , Santiago Martin Grieve , Stephan van der Zwaard , Yi Qin , Folkert W. Asselbergs , Tatiana Kuznetsova , Gisela Teixido-Tura , Alex P.W. Lee , Xiaomeng Li , Karim Lekadir , Laura Galian-Gay , Jose Rodriguez-Palomares , Ignacio Ferreira-Gonzalez , Lydia Dux-Santoy , Andrea Guala medRxiv 2026.11.19.26368489; doi: https://doi.org/10.64898/2026.11.19.26368489 Share This Article: Copy Citation Tools Automated echocardiography image quality assessment and its association with AI performance Pere Lopez-Gutierrez , Juan Garrido-Oliver , Mireia Bragulat-Arevalo , Juan Pablo Betancur-Rengifo , Nicholas Cauwenberghs , Alberto Morales-Galan , Marawan Elbatel , Machteld J. Boonstra , Francesca Ferrando , Horacio Majul , Lisandro Rivas Catoni , Santiago Martin Grieve , Stephan van der Zwaard , Yi Qin , Folkert W. Asselbergs , Tatiana Kuznetsova , Gisela Teixido-Tura , Alex P.W. Lee , Xiaomeng Li , Karim Lekadir , Laura Galian-Gay , Jose Rodriguez-Palomares , Ignacio Ferreira-Gonzalez , Lydia Dux-Santoy , Andrea Guala medRxiv 2026.11.19.26368489; doi: https://doi.org/10.64898/2026.11.19.26368489 Citation Manager Formats BibTeX Bookends EasyBib EndNote (tagged) EndNote 8 (xml) Medlars Mendeley Papers RefWorks Tagged Ref Manager RIS Zotero Tweet Widget Facebook Like Google Plus One Subject Area Cardiovascular Medicine Subject Areas All Articles Addiction Medicine (642) Allergy and Immunology (928) Anesthesia (355) Cardiovascular Medicine (5060) Dentistry and Oral Medicine (496) Dermatology (433) Emergency Medicine (681) Endocrinology (including Diabetes Mellitus and Metabolic Disease) (1704) Epidemiology (16353) Forensic Medicine (37) Gastroenterology (1250) Genetic and Genomic Medicine (7334) Geriatric Medicine (761) Health Economics (1101) Health Informatics (5352) Health Policy (1480) Health Systems and Quality Improvement (1847) Hematology (604) HIV/AIDS (1389) Infectious Diseases (except HIV/AIDS) (16537) Intensive Care and Critical Care Medicine (1210) Medical Education (702) Medical Ethics (157) Nephrology (757) Neurology (7669) Nursing (380) Nutrition (1144) Obstetrics and Gynecology (1308) Occupational and Environmental Health (1046) Oncology (3769) Ophthalmology (1105) Orthopedics (414) Otolaryngology (474) Pain Medicine (508) Palliative Medicine (144) Pathology (727) Pediatrics (1877) Pharmacology and Therapeutics (759) Primary Care Research (793) Psychiatry and Clinical Psychology (6187) Public and Global Health (10051) Radiology and Imaging (2562) Rehabilitation Medicine and Physical Therapy (1540) Respiratory Medicine (1280) Rheumatology (666) Sexual and Reproductive Health (814) Sports Medicine (601) Surgery (809) Toxicology (109) Transplantation (321) Urology (309)
## Related Clinical Research

- [Cardiac Structure and Function in Afro‑Caribbeans in Tobago: Population Cohort Echocardiography Fi](https://medichelpline.com/clinical-feed/medrxiv-5-epidemiology-of-cardiac-structure-and-function-among-african-caribbeans-in.md)
- [Clinician gaze as a supervisory signal for ECG deep learning: matched-control proof-of-concept](https://medichelpline.com/clinical-feed/medrxiv-12-clinician-gaze-as-a-supervisory-signal-for-electrocardiogram-deep-learning-a.md)
- [Smartwatch AI approximates office blood-pressure screening for undiagnosed hypertension](https://medichelpline.com/clinical-feed/nature-0-harnessing-smartwatches-to-spot-undiagnosed-hypertension.md)
- [Cardiologist questions STREAM trial findings on deprescribing statins in older adults](https://medichelpline.com/clinical-feed/stat-news-1-opinion-stat-i-m-a-cardiologist-i-m-a-little-puzzled-about-the-blockbuster-new.md)
- [Understanding the CAC Test for Cardiovascular Risk Assessment](https://medichelpline.com/clinical-feed/aha-news-0-what-is-a-cac-test-and-should-you-get-one.md)

## Navigation
- [← Back to Cardiology Feed](https://medichelpline.com/clinical-feed/cardiology.md)
- [← All Clinical Specialties](https://medichelpline.com/clinical-feed.md)
## Medical & Regulatory Disclaimer

> [!CAUTION]
> MedicHelpline content is structured for research, educational, and professional discovery purposes. It does not constitute individual medical advice, clinical diagnosis, or treatment recommendations.
> Always verify dosing, contraindications, and regulatory alerts against official product labeling and primary regulatory sources before clinical decision-making.