EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC 4.0

Deep survival learning for prognosis prediction in non-metastatic castration-resistant prostate cancer

Chunyang Li, Julia Bohman iD, Vikas Patil, Richard McShinsky, Christina Yong, Zach Burningham, Ahmad Halwani

Health Informatics Journal · 2026-04 · 32 (2)

DOI: 10.1177/14604582261456059 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Health Informatics Journal, online erschienen am 28.05.2026. Verfasst von Chunyang Li und 6 weiteren Autor:innen. Beteiligte Einrichtung: VA Salt Lake City Health Care System, Salt Lake City, UT, USA. Hauptthemen laut MEDLINE: Prostatic Neoplasms, Castration-Resistant, Deep Learning. Die Arbeit nennt 38 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Background

Non-metastatic, castration-resistant prostate cancer (nmCRPC) is an advanced state of prostate cancer with variable prognosis; early identification of patient risk is crucial, so that clinicians can recommend optimal treatment.

Objective

Compare predictive models in identifying patient risk; evaluate the value of electronic healthcare record (EHR) time-series (TS) information in prediction.

Methods

We evaluated SurvTRACE, Weibull Time to Event Recurrent Neural Network (WTTE-RNN), and traditional Cox proportional hazards (CPH) models’ performance on EHR data from 12,819 nmCRPC patients in the Veterans Health Administration, using area under the receiver operating characteristic curve and Brier score.

Results

WTTE-RNN, which intrinsically uses EHR TS information, outperformed the other models without TS information. Feature-engineered TS information improved performances of CPH and especially SurvTRACE; with TS information, SurvTRACE outperformed WTTE-RNN.

Conclusion

Deep learning methods, whether intrinsically able to handle TS data or enhanced with TS information, can outperform traditional survival analysis in predicting risk.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Prostatic Neoplasms, Castration-ResistantDeep LearningHumansMalePrognosisAgedProportional Hazards ModelsPredictive Learning ModelsSurvival AnalysisPrediction AlgorithmsROC CurveUnited StatesElectronic Health RecordsMiddle Aged

Stichwörter der Autor:innen

castration-resistant prostate cancerdeep learningmachine learningnon-metastaticrisk predictionsurvival analysis

Autor:innen und Institutionen

  • VA Salt Lake City Health Care System, Salt Lake City, UT, USA.
  • Hematology and Hematologic Malignancy, School of Medicine, University of Utah, Salt Lake City, UT, USA.
  • Huntsman Cancer Institute, Salt Lake City, UT, USA.

ORCID-Profile: Julia Bohman

Zitieren

APA

Li, C., Bohman, J., Patil, V., McShinsky, R., Yong, C., Burningham, Z., & Halwani, A. (2026). Deep survival learning for prognosis prediction in non-metastatic castration-resistant prostate cancer. Health Informatics Journal, 32(2). https://doi.org/10.1177/14604582261456059

Vancouver

Li C, Bohman J, Patil V, McShinsky R, Yong C, Burningham Z, et al. Deep survival learning for prognosis prediction in non-metastatic castration-resistant prostate cancer. Health Informatics Journal. 2026;32(2). doi:10.1177/14604582261456059

BibTeX
@article{Li2026,
  author = {Li, Chunyang and Bohman, Julia and Patil, Vikas and McShinsky, Richard and Yong, Christina and Burningham, Zach and Halwani, Ahmad},
  title = {Deep survival learning for prognosis prediction in non-metastatic castration-resistant prostate cancer},
  journal = {Health Informatics Journal},
  year = {2026},
  volume = {32},
  number = {2},
  doi = {10.1177/14604582261456059}
}
Kostenlos starten →