A transfer learning-based multimodal model for early prediction of 90-day respiratory failure in dermatomyositis-associated interstitial lung disease - Takeaways - MDSpire

A transfer learning-based multimodal model for early prediction of 90-day respiratory failure in dermatomyositis-associated interstitial lung disease

  • By

  • Lihui Guo

  • Yaning Yao

  • Qirui Wu

  • Hui Wang

  • Caiyun Niu

  • Gang Wang

  • Fei Chen

  • July 16, 2026

Share

  • 1

    Dermatomyositis-associated interstitial lung disease (DM-ILD) can lead to respiratory failure and high mortality, especially in anti-MDA5 antibody-positive patients.

  • 2

    A multimodal model was developed to predict 90-day respiratory failure in DM-ILD patients using data collected within the first 48 hours of admission.

  • 3

    The model demonstrated an area under the receiver operating characteristic curve (AUC) of 0.967, indicating strong discriminative performance.

  • 4

    Key predictors of respiratory failure included arthritis, pulmonary function indices, laboratory markers, and latent CT features.

  • 5

    The study highlights the potential of using early admission data for risk assessment in DM-ILD when anti-MDA5 antibody results are delayed.

Original Source(s)

A transfer learning-based multimodal model for early prediction of 90-day respiratory failure in dermatomyositis-associated interstitial lung disease

  • by Lihui Guo, Yaning Yao, Qirui Wu, Hui Wang, Caiyun Niu, Gang Wang, Fei Chen

  • July 16, 2026

Related Content

Determinants of exercise capacity in heart failure with preserved ejection fraction: central hemodynamics, ventilatory efficiency, and peripheral muscle oxygen utilization and function

Editorial: Advancements in mechanical ventilation: understanding physiology to mitigate complications

Development and validation of a nomogram based on LASSO regression for predicting early postoperative polyp recurrence in patients with chronic rhinosinusitis with nasal polyps