Evaluating machine learning pipelines for multimodal neuroimaging in small cohorts: an ALS case study
Advancements in machine learning hold great promise for the analysis of multimodal neuroimaging data. They can help identify biomarkers and improve diagnosis for various neurological disorders. However, the application of such techniques for rare and heterogeneous diseases remains challenging due to...
Saved in:
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2025-06-01
|
| Series: | Frontiers in Neuroinformatics |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fninf.2025.1568116/full |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850159619657170944 |
|---|---|
| author | Shailesh Appukuttan Shailesh Appukuttan Aude-Marie Grapperon Aude-Marie Grapperon Mounir Mohamed El Mendili Hugo Dary Maxime Guye Annie Verschueren Jean-Philippe Ranjeva Shahram Attarian Wafaa Zaaraoui Matthieu Gilson |
| author_facet | Shailesh Appukuttan Shailesh Appukuttan Aude-Marie Grapperon Aude-Marie Grapperon Mounir Mohamed El Mendili Hugo Dary Maxime Guye Annie Verschueren Jean-Philippe Ranjeva Shahram Attarian Wafaa Zaaraoui Matthieu Gilson |
| author_sort | Shailesh Appukuttan |
| collection | DOAJ |
| description | Advancements in machine learning hold great promise for the analysis of multimodal neuroimaging data. They can help identify biomarkers and improve diagnosis for various neurological disorders. However, the application of such techniques for rare and heterogeneous diseases remains challenging due to small-cohorts available for acquiring data. Efforts are therefore commonly directed toward improving the classification models, in an effort to optimize outcomes given the limited data. In this study, we systematically evaluated the impact of various machine learning pipeline configurations, including scaling methods, feature selection, dimensionality reduction, and hyperparameter optimization. The efficacy of such components in the pipeline was evaluated on classification performance using multimodal MRI data from a cohort of 16 ALS patients and 14 healthy controls. Our findings reveal that, while certain pipeline components, such as subject-wise feature normalization, help improve classification outcomes, the overall influence of pipeline refinements on performance is modest. Feature selection and dimensionality reduction steps were found to have limited utility, and the choice of hyperparameter optimization strategies produced only marginal gains. Our results suggest that, for small-cohort studies, the emphasis should shift from extensive tuning of these pipelines to addressing data-related limitations, such as progressively expanding cohort size, integrating additional modalities, and maximizing the information extracted from existing datasets. This study provides a methodological framework to guide future research and emphasizes the need for dataset enrichment to improve clinical utility. |
| format | Article |
| id | doaj-art-81f8a6985b174aee9f8fe2bec32e1f79 |
| institution | OA Journals |
| issn | 1662-5196 |
| language | English |
| publishDate | 2025-06-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Neuroinformatics |
| spelling | doaj-art-81f8a6985b174aee9f8fe2bec32e1f792025-08-20T02:23:28ZengFrontiers Media S.A.Frontiers in Neuroinformatics1662-51962025-06-011910.3389/fninf.2025.15681161568116Evaluating machine learning pipelines for multimodal neuroimaging in small cohorts: an ALS case studyShailesh Appukuttan0Shailesh Appukuttan1Aude-Marie Grapperon2Aude-Marie Grapperon3Mounir Mohamed El Mendili4Hugo Dary5Maxime Guye6Annie Verschueren7Jean-Philippe Ranjeva8Shahram Attarian9Wafaa Zaaraoui10Matthieu Gilson11Aix Marseille Univ, CNRS, CRMBM, Marseille, FranceAix Marseille Univ, CNRS, INT, Marseille, FranceAix Marseille Univ, CNRS, CRMBM, Marseille, FranceAPHM, Hopital de la Timone, Referral Centre for Neuromuscular Diseases and ALS, Marseille, FranceAix Marseille Univ, CNRS, CRMBM, Marseille, FranceAix Marseille Univ, CNRS, CRMBM, Marseille, FranceAix Marseille Univ, CNRS, CRMBM, Marseille, FranceAPHM, Hopital de la Timone, Referral Centre for Neuromuscular Diseases and ALS, Marseille, FranceAix Marseille Univ, CNRS, CRMBM, Marseille, FranceAPHM, Hopital de la Timone, Referral Centre for Neuromuscular Diseases and ALS, Marseille, FranceAix Marseille Univ, CNRS, CRMBM, Marseille, FranceAix Marseille Univ, CNRS, INT, Marseille, FranceAdvancements in machine learning hold great promise for the analysis of multimodal neuroimaging data. They can help identify biomarkers and improve diagnosis for various neurological disorders. However, the application of such techniques for rare and heterogeneous diseases remains challenging due to small-cohorts available for acquiring data. Efforts are therefore commonly directed toward improving the classification models, in an effort to optimize outcomes given the limited data. In this study, we systematically evaluated the impact of various machine learning pipeline configurations, including scaling methods, feature selection, dimensionality reduction, and hyperparameter optimization. The efficacy of such components in the pipeline was evaluated on classification performance using multimodal MRI data from a cohort of 16 ALS patients and 14 healthy controls. Our findings reveal that, while certain pipeline components, such as subject-wise feature normalization, help improve classification outcomes, the overall influence of pipeline refinements on performance is modest. Feature selection and dimensionality reduction steps were found to have limited utility, and the choice of hyperparameter optimization strategies produced only marginal gains. Our results suggest that, for small-cohort studies, the emphasis should shift from extensive tuning of these pipelines to addressing data-related limitations, such as progressively expanding cohort size, integrating additional modalities, and maximizing the information extracted from existing datasets. This study provides a methodological framework to guide future research and emphasizes the need for dataset enrichment to improve clinical utility.https://www.frontiersin.org/articles/10.3389/fninf.2025.1568116/fullamyotrophic lateral sclerosismachine learningmultimodal MRIsmall cohortclassificationpipeline optimization |
| spellingShingle | Shailesh Appukuttan Shailesh Appukuttan Aude-Marie Grapperon Aude-Marie Grapperon Mounir Mohamed El Mendili Hugo Dary Maxime Guye Annie Verschueren Jean-Philippe Ranjeva Shahram Attarian Wafaa Zaaraoui Matthieu Gilson Evaluating machine learning pipelines for multimodal neuroimaging in small cohorts: an ALS case study Frontiers in Neuroinformatics amyotrophic lateral sclerosis machine learning multimodal MRI small cohort classification pipeline optimization |
| title | Evaluating machine learning pipelines for multimodal neuroimaging in small cohorts: an ALS case study |
| title_full | Evaluating machine learning pipelines for multimodal neuroimaging in small cohorts: an ALS case study |
| title_fullStr | Evaluating machine learning pipelines for multimodal neuroimaging in small cohorts: an ALS case study |
| title_full_unstemmed | Evaluating machine learning pipelines for multimodal neuroimaging in small cohorts: an ALS case study |
| title_short | Evaluating machine learning pipelines for multimodal neuroimaging in small cohorts: an ALS case study |
| title_sort | evaluating machine learning pipelines for multimodal neuroimaging in small cohorts an als case study |
| topic | amyotrophic lateral sclerosis machine learning multimodal MRI small cohort classification pipeline optimization |
| url | https://www.frontiersin.org/articles/10.3389/fninf.2025.1568116/full |
| work_keys_str_mv | AT shaileshappukuttan evaluatingmachinelearningpipelinesformultimodalneuroimaginginsmallcohortsanalscasestudy AT shaileshappukuttan evaluatingmachinelearningpipelinesformultimodalneuroimaginginsmallcohortsanalscasestudy AT audemariegrapperon evaluatingmachinelearningpipelinesformultimodalneuroimaginginsmallcohortsanalscasestudy AT audemariegrapperon evaluatingmachinelearningpipelinesformultimodalneuroimaginginsmallcohortsanalscasestudy AT mounirmohamedelmendili evaluatingmachinelearningpipelinesformultimodalneuroimaginginsmallcohortsanalscasestudy AT hugodary evaluatingmachinelearningpipelinesformultimodalneuroimaginginsmallcohortsanalscasestudy AT maximeguye evaluatingmachinelearningpipelinesformultimodalneuroimaginginsmallcohortsanalscasestudy AT annieverschueren evaluatingmachinelearningpipelinesformultimodalneuroimaginginsmallcohortsanalscasestudy AT jeanphilipperanjeva evaluatingmachinelearningpipelinesformultimodalneuroimaginginsmallcohortsanalscasestudy AT shahramattarian evaluatingmachinelearningpipelinesformultimodalneuroimaginginsmallcohortsanalscasestudy AT wafaazaaraoui evaluatingmachinelearningpipelinesformultimodalneuroimaginginsmallcohortsanalscasestudy AT matthieugilson evaluatingmachinelearningpipelinesformultimodalneuroimaginginsmallcohortsanalscasestudy |