Assessing optimal methods for transferring machine learning models to low-volume and imbalanced clinical datasets: experiences from predicting outcomes of Danish trauma patients.

Journal: Frontiers in digital health
Published Date:

Abstract

INTRODUCTION: Accurately predicting patient outcomes is crucial for improving healthcare delivery, but large-scale risk prediction models are often developed and tested on specific datasets where clinical parameters and outcomes may not fully reflect local clinical settings. Where this is the case, whether to opt for de-novo training of prediction models on local datasets, direct porting of externally trained models, or a transfer learning approach is not well studied, and constitutes the focus of this study. Using the clinical challenge of predicting mortality and hospital length of stay on a Danish trauma dataset, we hypothesized that a transfer learning approach of models trained on large external datasets would provide optimal prediction results compared to de-novo training on sparse but local datasets or directly porting externally trained models.

Authors

  • Andreas Skov Millarch
    Department of Organ Surgery and Transplantation, Copenhagen University Hospital, Rigshospitalet, Copenhagen, Denmark.
  • Alexander Bonde
    Department of Organ Surgery and Transplantation, Copenhagen University Hospital, Rigshospitalet, Copenhagen, Denmark.
  • Mikkel Bonde
    Department of Organ Surgery and Transplantation, Copenhagen University Hospital, Rigshospitalet, Copenhagen, Denmark.
  • Kiril Vadomovic Klein
    Department of Computer Science, University of Copenhagen, Copenhagen, Denmark.
  • Fredrik Folke
    Copenhagen Emergency Medical Services, University of Copenhagen, Ballerup, Denmark.
  • Søren Steemann Rudolph
    Department of Anesthesia, Center of Head and Orthopedics, Rigshospitalet, Copenhagen, Denmark.
  • Martin Sillesen
    Department of Organ Surgery and Transplantation, Copenhagen University Hospital, Rigshospitalet, Copenhagen, Denmark.

Keywords

No keywords available for this article.