A Deep Learning Model for Inferring Elevated Pulmonary Capillary Wedge Pressures From the 12-Lead Electrocardiogram

• Published in: JACC. Advances (Online) Vol. 1; no. 1; p. 100003
• Main Authors: Schlesinger, Daphne E., Diamant, Nathaniel, Raghu, Aniruddh, Reinertsen, Erik, Young, Katherine, Batra, Puneet, Pomerantsev, Eugene, Stultz, Collin M.
• Format: Journal Article
• Language: English
• Published: United States Elsevier Inc 01.03.2022; Elsevier
• Subjects: Cardiovascular; deep learning; ECG; Original Research; pulmonary artery occlusion pressure; pulmonary artery wedge pressure; pulmonary capillary wedge pressure; pulmonary artery wedge pressure; deep learning; pulmonary artery occlusion pressure; ECG; pulmonary capillary wedge pressure; mPCWP; RHC; HF; heart failure; electrocardiogram; mean pulmonary capillary wedge pressure; right heart catheterization
• ISSN: 2772-963X; 2772-963X
• DOI: 10.1016/j.jacadv.2022.100003

Central hemodynamic parameters are typically measured via pulmonary artery catherization—an invasive procedure that involves some risk to the patient and is not routinely available in all settings. This study sought to develop a noninvasive method to identify elevated mean pulmonary capillary wedge pressure (mPCWP). We leveraged data from 248,955 clinical records at the Massachusetts General Hospital to develop a deep learning model that can infer when the mPCWP >15 mmHg using the 12-lead electrocardiogram (ECG). Of these data, 242,216 records were used to pre-train a model that generates useful ECG representations. The remaining 6,739 records contain encounters with direct measurements of the mPCWP. Eighty percent of these data were used for model development and testing (5,390), and the remaining records comprise a holdout set (1,349) that was used to evaluate the model. We developed an associated unreliability score that identifies when model predictions are likely to be untrustworthy. The model achieves an area under the receiver operating characteristic curve (AUC) of 0.80 ± 0.02 (test set) and 0.79 ± 0.01 (holdout set). Model performance varies as a function of the unreliability, where patients with high unreliability scores correspond to a subgroup where model performance is poor: for example, patients in the holdout set with unreliability scores in the highest decile have a reduced AUC of 0.70 ± 0.06. The mPCWP can be inferred from the ECG, and the reliability of this inference can be measured. When invasive monitoring cannot be expeditiously performed, deep learning models may provide information that can inform clinical care. [Display omitted] Noninvasive measurement of patient hemodynamics remains an important goal in cardiovascular medicine. In this study, we developed a deep learning model that can infer elevated mean pulmonary capillary wedge pressure (mPCWP) from the 12-lead electrocardiogram. The model, Right Heart Catheterization Network (RHCNet), also identifies those predictions that are likely to correspond to misleading results, thereby helping clinical providers gauge when model predictions are untrustworthy. Our method may form the foundation for an effective tool for ruling out elevated left-sided filling pressures in selected patients. RHCNet is generally available at https://github.com/daphneschles/RHCnet and may noninvasively screen for an elevated mPCWP when invasive hemodynamic monitoring cannot be routinely performed.