Machine learning for medical imaging: methodological failures and recommendations for the future
Abstract Research in computer analysis of medical images bears many promises to improve patients’ health. However, a number of systematic challenges are slowing down the progress of the field, from limitations of the data, such as biases, to research incentives, such as optimizing for publication. I...
Main Authors: | , |
---|---|
Format: | Article |
Language: | English |
Published: |
Nature Portfolio
2022-04-01
|
Series: | npj Digital Medicine |
Online Access: | https://doi.org/10.1038/s41746-022-00592-y |
_version_ | 1797429225982001152 |
---|---|
author | Gaël Varoquaux Veronika Cheplygina |
author_facet | Gaël Varoquaux Veronika Cheplygina |
author_sort | Gaël Varoquaux |
collection | DOAJ |
description | Abstract Research in computer analysis of medical images bears many promises to improve patients’ health. However, a number of systematic challenges are slowing down the progress of the field, from limitations of the data, such as biases, to research incentives, such as optimizing for publication. In this paper we review roadblocks to developing and assessing methods. Building our analysis on evidence from the literature and data challenges, we show that at every step, potential biases can creep in. On a positive note, we also discuss on-going efforts to counteract these problems. Finally we provide recommendations on how to further address these problems in the future. |
first_indexed | 2024-03-09T09:10:08Z |
format | Article |
id | doaj.art-99904cc3fb274e63b731b3a72b494b9e |
institution | Directory Open Access Journal |
issn | 2398-6352 |
language | English |
last_indexed | 2024-03-09T09:10:08Z |
publishDate | 2022-04-01 |
publisher | Nature Portfolio |
record_format | Article |
series | npj Digital Medicine |
spelling | doaj.art-99904cc3fb274e63b731b3a72b494b9e2023-12-02T08:55:00ZengNature Portfolionpj Digital Medicine2398-63522022-04-01511810.1038/s41746-022-00592-yMachine learning for medical imaging: methodological failures and recommendations for the futureGaël Varoquaux0Veronika Cheplygina1INRIAIT University of CopenhagenAbstract Research in computer analysis of medical images bears many promises to improve patients’ health. However, a number of systematic challenges are slowing down the progress of the field, from limitations of the data, such as biases, to research incentives, such as optimizing for publication. In this paper we review roadblocks to developing and assessing methods. Building our analysis on evidence from the literature and data challenges, we show that at every step, potential biases can creep in. On a positive note, we also discuss on-going efforts to counteract these problems. Finally we provide recommendations on how to further address these problems in the future.https://doi.org/10.1038/s41746-022-00592-y |
spellingShingle | Gaël Varoquaux Veronika Cheplygina Machine learning for medical imaging: methodological failures and recommendations for the future npj Digital Medicine |
title | Machine learning for medical imaging: methodological failures and recommendations for the future |
title_full | Machine learning for medical imaging: methodological failures and recommendations for the future |
title_fullStr | Machine learning for medical imaging: methodological failures and recommendations for the future |
title_full_unstemmed | Machine learning for medical imaging: methodological failures and recommendations for the future |
title_short | Machine learning for medical imaging: methodological failures and recommendations for the future |
title_sort | machine learning for medical imaging methodological failures and recommendations for the future |
url | https://doi.org/10.1038/s41746-022-00592-y |
work_keys_str_mv | AT gaelvaroquaux machinelearningformedicalimagingmethodologicalfailuresandrecommendationsforthefuture AT veronikacheplygina machinelearningformedicalimagingmethodologicalfailuresandrecommendationsforthefuture |