AI-Powered Resting 12-Lead Electrocardiogram Algorithm for Predicting Low Peak Oxygen Consumption: Development and Validation Study - Report - MDSpire

AI-Powered Resting 12-Lead Electrocardiogram Algorithm for Predicting Low Peak Oxygen Consumption: Development and Validation Study

  • By

  • Shu-Chun Huang

  • Tieh-Cheng Fu

  • Michelle Liou

  • Yu-Chieh Huang

  • Sing-Ya Chang

  • Guan-Yi Huang

  • Hong-Ren Su

  • June 11, 2026

  • 0 min

Share

Clinical Report: AI-Enhanced ECG Algorithm for Forecasting Low Peak Oxygen Uptake

Overview

This study developed and validated an AI-based algorithm utilizing resting 12-lead ECG data to predict low peak oxygen uptake (peak V̇O2) in patients.

Background

Low peak V̇O2 is a significant predictor of cardiovascular disease and mortality, making its early identification crucial for patient management. Traditional assessment through cardiopulmonary exercise testing (CPET) is resource-intensive and not widely implemented in clinical practice.

Data Highlights

DatasetSample SizePeak V̇O2 Threshold
Training965<14 mL/kg/min
Validation242<14 mL/kg/min

Key Findings

  • The AI algorithm was trained on ECG data from 965 patients and validated on 242 patients.
  • Low peak V̇O2 was defined as a peak V̇O2 of less than 14 mL/kg/min.
  • Resting 12-lead ECG signals were utilized to estimate low peak V̇O2 effectively.
  • The study included a higher proportion of patients with cardiopulmonary diseases compared to previous studies.
  • ECG abnormalities correlate with decreased cardiorespiratory fitness.

Clinical Implications

The AI-enhanced ECG algorithm may facilitate earlier identification of patients at risk for low peak V̇O2.

Conclusion

The development of an AI-based ECG algorithm presents a new approach to screening for low peak V̇O2.

Related Resources & Content

  1. Prokopidis et al., ESC Heart Failure, 2025 -- Prognostic impact of peak oxygen consumption in heart failure: A systematic review and meta‐analysis
  2. JACC: Advances, 2026 -- Prediction of Pregnancy-Related Cardiovascular Outcomes Using Electrocardiogram-Based Deep Learning Estimation of Cardiorespiratory Fitness
  3. Frontiers in Cardiovascular Medicine — Artificial intelligence applied to post-resuscitation ECGs for early prognostication after out-of-hospital cardiac arrest
  4. npj Digital Medicine — Artificial intelligence-enabled analysis of handheld single-lead electrocardiograms to predict incident atrial fibrillation: an analysis of the VITAL-AF randomized trial
  5. npj Digital Medicine — Electrocardiographic Age from Wearable Devices and Its Link to Atrial Fibrillation
  6. Pediatric Cardiology — AI-Driven ECG Analysis for Determining Age and Gender in Pediatric Populations
  7. Artificial intelligence applied to post-resuscitation ECGs for early prognostication after out-of-hospital cardiac arrest
  8. Artificial intelligence-enabled analysis of handheld single-lead electrocardiograms to predict incident atrial fibrillation: an analysis of the VITAL-AF randomized trial
  9. Electrocardiographic Age from Wearable Devices and Its Link to Atrial Fibrillation
  10. Prognostic impact of peak oxygen consumption in heart failure: A systematic review and meta‐analysis - Prokopidis - 2025 - ESC Heart Failure - Wiley Online Library
  11. Prediction of Pregnancy-Related Cardiovascular Outcomes Using Electrocardiogram-Based Deep Learning Estimation of Cardiorespiratory Fitness | JACC: Advances

Original Source(s)

AI-Powered Resting 12-Lead Electrocardiogram Algorithm for Predicting Low Peak Oxygen Consumption: Development and Validation Study

  • by Shu-Chun Huang, Tieh-Cheng Fu, Michelle Liou, Yu-Chieh Huang, Sing-Ya Chang, Guan-Yi Huang, Hong-Ren Su

  • June 11, 2026

Related Content

Investigational Drug Shows Potentialin Phase 3 Obesity Trial

In a 76-week randomized trial, patients with obesity without diabetes who received survodutide achieved greater weight loss and favorable changes in several metabolic measures compared with those who received placebo.

Pretreatment cardiometabolic risk in newly diagnosed breast cancer: a molecular subtype analysis

Divergent PERTs: A call for unity via harmonized training programs