Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

Longitudinal outcome prediction of prostate cancer patients on active surveillance using multiple instance learning

Winzell, Filip LU ; Arvidsson, Ida LU orcid ; Åström, Kalle LU orcid ; Overgaard, Niels Christian LU ; Marginean, Felicia Elena LU orcid ; Simoulis, Athanasios LU orcid ; Bjartell, Anders LU ; Krzyzanowska, Agnieszka LU and Heyden, Anders LU orcid (2025) In Journal of Medical Imaging 12(6).
Abstract

Purpose: To avoid over-treatment of prostate cancer patients following screening for elevated prostate-specific antigen levels, keeping patients on active surveillance has been suggested as an alternative to radical treatment. This means recurring visits for patients with low-grade cancer to monitor progression. Our aim was to develop an artificial intelligence-based model that can identify high-risk patients in a cohort of prostate cancer patients on active surveillance. Approach: We have developed a multiple instance learning-based framework for predicting the longitudinal outcomes for prostate cancer patients on active surveillance. Our models were trained only on whole-slide images with patient-level labels without using explicit... (More)

Purpose: To avoid over-treatment of prostate cancer patients following screening for elevated prostate-specific antigen levels, keeping patients on active surveillance has been suggested as an alternative to radical treatment. This means recurring visits for patients with low-grade cancer to monitor progression. Our aim was to develop an artificial intelligence-based model that can identify high-risk patients in a cohort of prostate cancer patients on active surveillance. Approach: We have developed a multiple instance learning-based framework for predicting the longitudinal outcomes for prostate cancer patients on active surveillance. Our models were trained only on whole-slide images with patient-level labels without using explicit Gleason grades. We employed the UNI-2 foundation model and the well-established attention-based multiple instance learning approach. We further evaluated our models by fitting Cox proportional hazards models and testing them on an external dataset. Results: With this approach, we achieved an average area under the receiver operator characteristic curve of 0.958 (95% CI, 0.957 to 0.959). Fitting Cox models to the predicted probabilities achieved a C-index of 0.824 and a hazard ratio of 2.32. However, all models showed a large drop in performance when evaluated on an external dataset. Conclusion: We show that avoiding Gleason grades is beneficial for longitudinal outcome prediction of prostate cancer. Our results suggest that benign prostate tissue contains prognostic information. However, before our models could be used clinically, much more work remains to improve the generalization.

(Less)
Please use this url to cite or link to this publication:
author
; ; ; ; ; ; ; and
organization
publishing date
type
Contribution to journal
publication status
published
subject
keywords
artificial intelligence, deep learning, digital pathology, multiple instance learning, patient outcome, prostate cancer, whole slide images
in
Journal of Medical Imaging
volume
12
issue
6
article number
061408
publisher
SPIE
external identifiers
  • scopus:105026343953
  • pmid:41098618
ISSN
2329-4302
DOI
10.1117/1.JMI.12.6.061408
language
English
LU publication?
yes
additional info
Publisher Copyright: © The Authors. Published by SPIE under a Creative Commons Attribution 4.0 International License. Distribution or reproduction of this work in whole or in part requires full attribution of the original publication, including its DOI.
id
49f20452-3528-486e-a039-ee0fba104e71
date added to LUP
2026-02-17 13:33:29
date last changed
2026-02-18 03:16:02
@article{49f20452-3528-486e-a039-ee0fba104e71,
  abstract     = {{<p>Purpose: To avoid over-treatment of prostate cancer patients following screening for elevated prostate-specific antigen levels, keeping patients on active surveillance has been suggested as an alternative to radical treatment. This means recurring visits for patients with low-grade cancer to monitor progression. Our aim was to develop an artificial intelligence-based model that can identify high-risk patients in a cohort of prostate cancer patients on active surveillance. Approach: We have developed a multiple instance learning-based framework for predicting the longitudinal outcomes for prostate cancer patients on active surveillance. Our models were trained only on whole-slide images with patient-level labels without using explicit Gleason grades. We employed the UNI-2 foundation model and the well-established attention-based multiple instance learning approach. We further evaluated our models by fitting Cox proportional hazards models and testing them on an external dataset. Results: With this approach, we achieved an average area under the receiver operator characteristic curve of 0.958 (95% CI, 0.957 to 0.959). Fitting Cox models to the predicted probabilities achieved a C-index of 0.824 and a hazard ratio of 2.32. However, all models showed a large drop in performance when evaluated on an external dataset. Conclusion: We show that avoiding Gleason grades is beneficial for longitudinal outcome prediction of prostate cancer. Our results suggest that benign prostate tissue contains prognostic information. However, before our models could be used clinically, much more work remains to improve the generalization.</p>}},
  author       = {{Winzell, Filip and Arvidsson, Ida and Åström, Kalle and Overgaard, Niels Christian and Marginean, Felicia Elena and Simoulis, Athanasios and Bjartell, Anders and Krzyzanowska, Agnieszka and Heyden, Anders}},
  issn         = {{2329-4302}},
  keywords     = {{artificial intelligence; deep learning; digital pathology; multiple instance learning; patient outcome; prostate cancer; whole slide images}},
  language     = {{eng}},
  month        = {{10}},
  number       = {{6}},
  publisher    = {{SPIE}},
  series       = {{Journal of Medical Imaging}},
  title        = {{Longitudinal outcome prediction of prostate cancer patients on active surveillance using multiple instance learning}},
  url          = {{http://dx.doi.org/10.1117/1.JMI.12.6.061408}},
  doi          = {{10.1117/1.JMI.12.6.061408}},
  volume       = {{12}},
  year         = {{2025}},
}