Population heterogeneity in clinical cohorts affects the predictive accuracy of brain imaging.
Brain imaging research enjoys increasing adoption of supervised machine learning for single-participant disease classification. Yet, the success of these algorithms likely depends on population diversity, including demographic differences and other factors that may be outside of primary scientific i...
| Main Authors: | , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Public Library of Science (PLoS)
2022-04-01
|
| Series: | PLoS Biology |
| Online Access: | https://doi.org/10.1371/journal.pbio.3001627 |
| _version_ | 1828746468363599872 |
|---|---|
| author | Oualid Benkarim Casey Paquola Bo-Yong Park Valeria Kebets Seok-Jun Hong Reinder Vos de Wael Shaoshi Zhang B T Thomas Yeo Michael Eickenberg Tian Ge Jean-Baptiste Poline Boris C Bernhardt Danilo Bzdok |
| author_facet | Oualid Benkarim Casey Paquola Bo-Yong Park Valeria Kebets Seok-Jun Hong Reinder Vos de Wael Shaoshi Zhang B T Thomas Yeo Michael Eickenberg Tian Ge Jean-Baptiste Poline Boris C Bernhardt Danilo Bzdok |
| author_sort | Oualid Benkarim |
| collection | DOAJ |
| description | Brain imaging research enjoys increasing adoption of supervised machine learning for single-participant disease classification. Yet, the success of these algorithms likely depends on population diversity, including demographic differences and other factors that may be outside of primary scientific interest. Here, we capitalize on propensity scores as a composite confound index to quantify diversity due to major sources of population variation. We delineate the impact of population heterogeneity on the predictive accuracy and pattern stability in 2 separate clinical cohorts: the Autism Brain Imaging Data Exchange (ABIDE, n = 297) and the Healthy Brain Network (HBN, n = 551). Across various analysis scenarios, our results uncover the extent to which cross-validated prediction performances are interlocked with diversity. The instability of extracted brain patterns attributable to diversity is located preferentially in regions part of the default mode network. Collectively, our findings highlight the limitations of prevailing deconfounding practices in mitigating the full consequences of population diversity. |
| first_indexed | 2024-04-14T04:25:10Z |
| format | Article |
| id | doaj.art-64f312f7b0244c1d91348ad7f49b43f9 |
| institution | Directory Open Access Journal |
| issn | 1544-9173 1545-7885 |
| language | English |
| last_indexed | 2024-04-14T04:25:10Z |
| publishDate | 2022-04-01 |
| publisher | Public Library of Science (PLoS) |
| record_format | Article |
| series | PLoS Biology |
| spelling | doaj.art-64f312f7b0244c1d91348ad7f49b43f92022-12-22T02:12:21ZengPublic Library of Science (PLoS)PLoS Biology1544-91731545-78852022-04-01204e300162710.1371/journal.pbio.3001627Population heterogeneity in clinical cohorts affects the predictive accuracy of brain imaging.Oualid BenkarimCasey PaquolaBo-Yong ParkValeria KebetsSeok-Jun HongReinder Vos de WaelShaoshi ZhangB T Thomas YeoMichael EickenbergTian GeJean-Baptiste PolineBoris C BernhardtDanilo BzdokBrain imaging research enjoys increasing adoption of supervised machine learning for single-participant disease classification. Yet, the success of these algorithms likely depends on population diversity, including demographic differences and other factors that may be outside of primary scientific interest. Here, we capitalize on propensity scores as a composite confound index to quantify diversity due to major sources of population variation. We delineate the impact of population heterogeneity on the predictive accuracy and pattern stability in 2 separate clinical cohorts: the Autism Brain Imaging Data Exchange (ABIDE, n = 297) and the Healthy Brain Network (HBN, n = 551). Across various analysis scenarios, our results uncover the extent to which cross-validated prediction performances are interlocked with diversity. The instability of extracted brain patterns attributable to diversity is located preferentially in regions part of the default mode network. Collectively, our findings highlight the limitations of prevailing deconfounding practices in mitigating the full consequences of population diversity.https://doi.org/10.1371/journal.pbio.3001627 |
| spellingShingle | Oualid Benkarim Casey Paquola Bo-Yong Park Valeria Kebets Seok-Jun Hong Reinder Vos de Wael Shaoshi Zhang B T Thomas Yeo Michael Eickenberg Tian Ge Jean-Baptiste Poline Boris C Bernhardt Danilo Bzdok Population heterogeneity in clinical cohorts affects the predictive accuracy of brain imaging. PLoS Biology |
| title | Population heterogeneity in clinical cohorts affects the predictive accuracy of brain imaging. |
| title_full | Population heterogeneity in clinical cohorts affects the predictive accuracy of brain imaging. |
| title_fullStr | Population heterogeneity in clinical cohorts affects the predictive accuracy of brain imaging. |
| title_full_unstemmed | Population heterogeneity in clinical cohorts affects the predictive accuracy of brain imaging. |
| title_short | Population heterogeneity in clinical cohorts affects the predictive accuracy of brain imaging. |
| title_sort | population heterogeneity in clinical cohorts affects the predictive accuracy of brain imaging |
| url | https://doi.org/10.1371/journal.pbio.3001627 |
| work_keys_str_mv | AT oualidbenkarim populationheterogeneityinclinicalcohortsaffectsthepredictiveaccuracyofbrainimaging AT caseypaquola populationheterogeneityinclinicalcohortsaffectsthepredictiveaccuracyofbrainimaging AT boyongpark populationheterogeneityinclinicalcohortsaffectsthepredictiveaccuracyofbrainimaging AT valeriakebets populationheterogeneityinclinicalcohortsaffectsthepredictiveaccuracyofbrainimaging AT seokjunhong populationheterogeneityinclinicalcohortsaffectsthepredictiveaccuracyofbrainimaging AT reindervosdewael populationheterogeneityinclinicalcohortsaffectsthepredictiveaccuracyofbrainimaging AT shaoshizhang populationheterogeneityinclinicalcohortsaffectsthepredictiveaccuracyofbrainimaging AT btthomasyeo populationheterogeneityinclinicalcohortsaffectsthepredictiveaccuracyofbrainimaging AT michaeleickenberg populationheterogeneityinclinicalcohortsaffectsthepredictiveaccuracyofbrainimaging AT tiange populationheterogeneityinclinicalcohortsaffectsthepredictiveaccuracyofbrainimaging AT jeanbaptistepoline populationheterogeneityinclinicalcohortsaffectsthepredictiveaccuracyofbrainimaging AT boriscbernhardt populationheterogeneityinclinicalcohortsaffectsthepredictiveaccuracyofbrainimaging AT danilobzdok populationheterogeneityinclinicalcohortsaffectsthepredictiveaccuracyofbrainimaging |