Cross-population evaluation of cervical cancer risk prediction algorithms.

Cervical cancer prevention External validation Hidden Markov model Matrix factorization Neural network Risk prediction

Journal

International journal of medical informatics
ISSN: 1872-8243
Titre abrégé: Int J Med Inform
Pays: Ireland
ID NLM: 9711057

Informations de publication

Date de publication:
Jan 2024
Historique:
received: 29 09 2023
revised: 11 10 2023
accepted: 12 11 2023
pubmed: 29 11 2023
medline: 29 11 2023
entrez: 28 11 2023
Statut: ppublish

Résumé

Cervical cancer is a preventable disease, despite being one of the most common types of female cancers worldwide. Integrating existing programs for cervical cancer screening with personalized risk prediction algorithms can improve population-level cancer prevention by enabling more targeted screening and contrive preventive healthcare innovations. While algorithms developed for cervical cancer risk prediction have shown promising performance in internal validation on more homogeneous populations, their ability to generalize to external populations remains to be assessed. To address this gap, we perform a cross-population comparative study of personalized prediction algorithms for more personalized cervical cancer screening. Using data from the Norwegian and Estonian populations, the algorithms are validated on internal and external datasets to study their potential biases and limitations when applied to different populations. We evaluate the algorithms in predicting progression from low-grade precancerous cervical lesions, simulating a clinically relevant application of more personalized risk stratification. As expected, our numerical experiments show that algorithm performance varies depending on the population. However, some algorithms show strong generalization capacity across different data sources. Using Kaplan-Meier estimates, we demonstrate the strengths and limitations of the algorithms in detecting cancer progression over time by comparing to the trends observed from data. We assess their overall discrimination performance in personalized risk predictions by analyzing the accuracy and confidence in individual risk estimates. This study examines the effectiveness of personalized prediction algorithms across different populations. Our results demonstrate the potential for generalizing risk prediction algorithms to external populations. These findings highlight the importance of considering population diversity when developing risk prediction algorithms.

Sections du résumé

BACKGROUND BACKGROUND
Cervical cancer is a preventable disease, despite being one of the most common types of female cancers worldwide. Integrating existing programs for cervical cancer screening with personalized risk prediction algorithms can improve population-level cancer prevention by enabling more targeted screening and contrive preventive healthcare innovations. While algorithms developed for cervical cancer risk prediction have shown promising performance in internal validation on more homogeneous populations, their ability to generalize to external populations remains to be assessed.
METHODS METHODS
To address this gap, we perform a cross-population comparative study of personalized prediction algorithms for more personalized cervical cancer screening. Using data from the Norwegian and Estonian populations, the algorithms are validated on internal and external datasets to study their potential biases and limitations when applied to different populations. We evaluate the algorithms in predicting progression from low-grade precancerous cervical lesions, simulating a clinically relevant application of more personalized risk stratification.
RESULTS RESULTS
As expected, our numerical experiments show that algorithm performance varies depending on the population. However, some algorithms show strong generalization capacity across different data sources. Using Kaplan-Meier estimates, we demonstrate the strengths and limitations of the algorithms in detecting cancer progression over time by comparing to the trends observed from data. We assess their overall discrimination performance in personalized risk predictions by analyzing the accuracy and confidence in individual risk estimates.
DISCUSSION AND CONCLUSION CONCLUSIONS
This study examines the effectiveness of personalized prediction algorithms across different populations. Our results demonstrate the potential for generalizing risk prediction algorithms to external populations. These findings highlight the importance of considering population diversity when developing risk prediction algorithms.

Identifiants

pubmed: 38016404
pii: S1386-5056(23)00315-5
doi: 10.1016/j.ijmedinf.2023.105297
pii:
doi:

Types de publication

Journal Article

Langues

eng

Sous-ensembles de citation

IM

Pagination

105297

Informations de copyright

Copyright © 2023 The Authors. Published by Elsevier B.V. All rights reserved.

Déclaration de conflit d'intérêts

Declaration of Competing Interest The authors declare that they have no competing interests.

Auteurs

Severin Elvatun (S)

Department of Registry Informatics, Cancer Registry of Norway, Ullernchausseen 64, 0379 Oslo, Norway. Electronic address: sela@kreftregisteret.no.

Daan Knoors (D)

Department of Registry Informatics, Cancer Registry of Norway, Norway.

Mari Nygård (M)

Department of Research, Cancer Registry of Norway, Norway.

Anneli Uusküla (A)

Department of Family Medicine and Public Health, University of Tartu, Estonia.

Andres Võrk (A)

Institute of Economics, University of Tartu, Estonia.

Jan F Nygård (JF)

Department of Registry Informatics, Cancer Registry of Norway, Department of Physics and Technology, UiT The Arctic University of Norway, Norway.

Classifications MeSH