Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Video

Updated: May 31, 2026

Lumped-Parameter and Finite Element Modeling of Heart Failure with Preserved Ejection Fraction
09:20

Lumped-Parameter and Finite Element Modeling of Heart Failure with Preserved Ejection Fraction

Published on: February 13, 2021

Explainable machine learning for estimation of elevated left ventricular filling pressure: a multicenter validation.

Yutaka Nakamura1, Nobuyuki Kagiyama2,3,4, Sirish Shrestha5

  • 1Department of Cardiovascular Biology and Medicine, Juntendo University Graduate School of Medicine, Tokyo, Japan.

Journal of Echocardiography
|May 29, 2026
PubMed
Summary

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Prognostic Impact of Biplane Vena Contracta Width in Patients with Atrial Functional Mitral Regurgitation.

European heart journal. Cardiovascular Imaging·2026
Same author

Echocardiography Finds an Unusual Cause of Atrial Fibrillation.

CASE (Philadelphia, Pa.)·2026
Same author

Early Outcomes Between the Two Latest Generation Mitral Valve Transcatheter Edge-to-Edge Repair Devices: OCEAN-Mitral Registry.

Catheterization and cardiovascular interventions : official journal of the Society for Cardiac Angiography & Interventions·2026
Same author

Cerebral Infarction Diagnosed as Isolated Cardiac Sarcoidosis.

NMC case report journal·2026
Same author

Cheyne-Stokes respiration screening using cyclic variation in heart rate detected by Holter electrocardiogram.

Journal of cardiology·2026
Same author

Comparison of the Expert Guidelines With Artificial Intelligence-Driven Echocardiographic Assessment of Diastolic Function.

medRxiv : the preprint server for health sciences·2026
This summary is machine-generated.

Explainable machine learning (ML) models accurately estimate left ventricular filling pressure (LVFP), outperforming guideline-recommended algorithms. These interpretable ML tools offer patient-level insights for clinical decision-making.

Area of Science:

  • Cardiology
  • Medical Informatics
  • Machine Learning

Background:

  • Guideline-recommended algorithms (GL-algorithm) often yield indeterminate left ventricular filling pressure (LVFP) results.
  • Machine learning (ML) methods offer high accuracy but lack clinical interpretability.
  • Explainable ML models are needed for reliable clinical application in estimating LVFP.

Purpose of the Study:

  • Develop an explainable ML model for LVFP estimation.
  • Provide patient-level interpretation using gold-standard right heart catheterization (RHC) data.
  • Compare the performance of explainable ML models against GL-algorithms.

Main Methods:

  • Retrospective enrollment of 956 patients undergoing echocardiography and RHC.
  • Training two extreme gradient boosting models to estimate elevated pulmonary artery wedge pressure (PAWP) as a surrogate for LVFP.
Keywords:
Explainable AIHeart failureLeft ventricular filling pressureMachine learningSHAP

More Related Videos

Cardiac Response to β-Adrenergic Stimulation Determined by Pressure-Volume Loop Analysis
08:05

Cardiac Response to β-Adrenergic Stimulation Determined by Pressure-Volume Loop Analysis

Published on: May 19, 2021

Related Experiment Videos

Last Updated: May 31, 2026

Lumped-Parameter and Finite Element Modeling of Heart Failure with Preserved Ejection Fraction
09:20

Lumped-Parameter and Finite Element Modeling of Heart Failure with Preserved Ejection Fraction

Published on: February 13, 2021

Cardiac Response to β-Adrenergic Stimulation Determined by Pressure-Volume Loop Analysis
08:05

Cardiac Response to β-Adrenergic Stimulation Determined by Pressure-Volume Loop Analysis

Published on: May 19, 2021

  • Model 1 used GL-algorithm variables; Model 2 used variables selected by Shapley additive explanations (SHAP) values.
  • Comparing model performance using area under the receiver-operating characteristic curve (AUROC) on external test data.

    • Main Results:

    • GL-algorithm classified 42.7% of patients as indeterminate for LVFP.
    • ML models achieved full patient classification.
    • Both ML models significantly outperformed GL-algorithm in estimating LVFP (AUROC 0.82-0.83 vs. 0.72).
    • SHAP force plots provided patient-level variable contribution insights.

    Conclusions:

    • Explainable ML models demonstrate superior performance over GL-algorithms for LVFP estimation.
    • The developed ML models offer enhanced accuracy and patient-level interpretability.
    • This approach provides a user-friendly tool for clinicians in assessing LVFP.