Multimodal fusion of EHR and ECG based on deep learning for predicting new-onset coronary heart disease in cancer patients - Takeaways - MDSpire

Multimodal fusion of EHR and ECG based on deep learning for predicting new-onset coronary heart disease in cancer patients

  • By

  • Sheng Zhang

  • Wei Wang

  • June 17, 2026

  • 0 min

Share

  • 1

    A multimodal deep learning model was developed to predict new-onset coronary heart disease (CHD) in cancer patients using EHR and ECG data.

  • 2

    The study included 1262 cancer patients, with 57.2% developing new-onset CHD during hospitalization or follow-up.

  • 3

    The CNN-LSTM model achieved an AUC of 0.975, outperforming seven conventional machine learning methods in predicting new-onset CHD.

  • 4

    Subgroup analysis showed the model yielded an AUC of 0.924 for early-stage patients and 0.888 for late-stage patients.

  • 5

    Key predictive features identified included age, CRP, cTnI, and ECG markers such as heart rate and QT intervals.

Original Source(s)

Multimodal fusion of EHR and ECG based on deep learning for predicting new-onset coronary heart disease in cancer patients

  • by Sheng Zhang, Wei Wang

  • June 17, 2026

Related Content

Liquid biopsy, multi-cancer early detection, and artificial intelligence: new frontiers in cancer screening from a technological and immunological perspective

Artificial intelligence–enabled liquid biopsy in cancer: a systematic review and meta- analysis of diagnostic performance and biological implications

Conversion-oriented multimodal therapy enabling definitive surgery in a patient with FIGO stage IVA cervical cancer at high risk of fistula formation: a case report