Predicting Cardiopulmonary Exercise Testing Performance in Patients Undergoing Transthoracic Echocardiography - An AI-Based, Multimodal Model

Background and Aims Cardiopulmonary exercise test (CPET) performance strongly correlates with survival in several disease states including heart failure, but has limited accessibility. Meanwhile, transthoracic echocardiography (TTE) is a far more widely available tool. We sought to develop an artificial intelligence (AI) algorithm using TTE and data from electronic medical records to predict CPET peak oxygen consumption (peak VO2) ≤ 14 mL/kg/min. Methods An AI model was trained to predict peak VO2 ≤14 mL/kg/min from TTE images, structured TTE reports, demographics, medications, labs, and vitals. The training set included patients with a TTE within 6 months of a CPET. Performance was retrospectively tested in a held-out group from the development cohort and an external regional validation cohort. Results 1,127 CPET studies paired with concomitant TTE were identified. The best performance was achieved by using all components (TTE images, all structured clinical data). The model performed well at predicting a peak VO2 ≤14 mL/kg/min, with an AUROC of 0.84 (development cohort) and 0.80 (external validation cohort). It performed consistently well using higher (≤18 mL/kg/min) and lower (≤12 mL/kg/min) cut-offs. Conclusions This multimodal AI model effectively categorized patients into low and high risk predicted peak VO2, demonstrating the potential to identify previously unrecognized high risk patients when CPET is not available.

Authors

Institutions

Publication Details

Journal
European Heart Journal - Digital Health
Published
2026-11-18
DOI
https://doi.org/10.1093/ehjdh/ztag172
Citations
2
Primary Topic
Cardiovascular and exercise physiology
Type
article
Field-Weighted Citation Impact
4.93
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Predicting Cardiopulmonary Exercise Testing Performance in Patients Undergoing Transthoracic Echocardiography - An AI-Based, Multimodal Model

Ashley Beecy, Chris R. Kelsey, Erica Lancet, Kevin J. Clerkin et al.
2 citations
European Heart Journal - Digital Health
Cardiovascular and exercise physiology
4.93
article

Predicting Cardiopulmonary Exercise Testing Performance in Patients Undergoing Transthoracic Echocardiography - An AI-Based, Multimodal Model

Ashley Beecy, Chris R. Kelsey, Erica Lancet, Kevin J. Clerkin, Volodymyr Kuleshov, Shudhanshu Alishetti, Deepa Kumaraiah, Deborah L. Estrin, Weishen Pan, Zhe Huang, David T. Majure, Jeffrey A. Ruhl, Fei Wang, Zhenzhen Liu, Gabriel T. Sayer, Nir Uriel, Rochelle L. Goldsmith, Naomi Tesfuzigta, David vanMaanan
article en
2 citations

Abstract

Background and Aims Cardiopulmonary exercise test (CPET) performance strongly correlates with survival in several disease states including heart failure, but has limited accessibility. Meanwhile, transthoracic echocardiography (TTE) is a far more widely available tool. We sought to develop an artificial intelligence (AI) algorithm using TTE and data from electronic medical records to predict CPET peak oxygen consumption (peak VO2) ≤ 14 mL/kg/min. Methods An AI model was trained to predict peak VO2 ≤14 mL/kg/min from TTE images, structured TTE reports, demographics, medications, labs, and vitals. The training set included patients with a TTE within 6 months of a CPET. Performance was retrospectively tested in a held-out group from the development cohort and an external regional validation cohort. Results 1,127 CPET studies paired with concomitant TTE were identified. The best performance was achieved by using all components (TTE images, all structured clinical data). The model performed well at predicting a peak VO2 ≤14 mL/kg/min, with an AUROC of 0.84 (development cohort) and 0.80 (external validation cohort). It performed consistently well using higher (≤18 mL/kg/min) and lower (≤12 mL/kg/min) cut-offs. Conclusions This multimodal AI model effectively categorized patients into low and high risk predicted peak VO2, demonstrating the potential to identify previously unrecognized high risk patients when CPET is not available.

European Heart Journal - Digital Health
NewYork–Presbyterian Hospital (US), NewYork–Presbyterian Brooklyn Methodist Hospital (US), Cornell University (US), Columbia University Irving Medical Center (US), New York Hospital Queens (US), Presbyterian Hospital (US), Weill Cornell Medicine (US), Cornell Tech (US), Weill Cornell Medical Center (US)
Good health and well-being
Openalex Percentile: Top 8%
Cardiovascular and exercise physiology
4.93
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon