Predicting disease severity in multiple sclerosis using multimodal data and machine learning.

Autor: Andorra M; Institut d'Investigacions Biomediques August Pi Sunyer (IDIBAPS) and Hospital Clinic Barcelona, Barcelona, Spain., Freire A; School of Management, Pompeu Fabra University, Barcelona, Spain.; UPF Barcelona School of Management, Balmes 132, 08008, Barcelona, Spain., Zubizarreta I; Institut d'Investigacions Biomediques August Pi Sunyer (IDIBAPS) and Hospital Clinic Barcelona, Barcelona, Spain., de Rosbo NK; Department of Neurosciences, Rehabilitation, Ophthalmology, Genetics, Maternal and Child Health, University of Genoa, Genoa, Italy.; IRCCS Ospedale Policlinico San Martino, Genoa, Italy., Bos SD; University of Oslo, Oslo, Norway.; Oslo University Hospital, Oslo, Norway., Rinas M; Institute for Computational Biomedicine, Heidelberg University Hospital, and Heidelberg University, Heidelberg, Germany., Høgestøl EA; University of Oslo, Oslo, Norway.; Oslo University Hospital, Oslo, Norway., de Rodez Benavent SA; University of Oslo, Oslo, Norway.; Oslo University Hospital, Oslo, Norway., Berge T; Oslo University Hospital, Oslo, Norway.; Oslo Metropolitan University, Oslo, Norway., Brune-Ingebretse S; University of Oslo, Oslo, Norway.; Oslo University Hospital, Oslo, Norway., Ivaldi F; Department of Internal Medicine, University of Genoa, Genoa, Italy., Cellerino M; Department of Neurosciences, Rehabilitation, Ophthalmology, Genetics, Maternal and Child Health, University of Genoa, Genoa, Italy., Pardini M; Department of Neurosciences, Rehabilitation, Ophthalmology, Genetics, Maternal and Child Health, University of Genoa, Genoa, Italy.; IRCCS Ospedale Policlinico San Martino, Genoa, Italy., Vila G; Institut d'Investigacions Biomediques August Pi Sunyer (IDIBAPS) and Hospital Clinic Barcelona, Barcelona, Spain., Pulido-Valdeolivas I; Institut d'Investigacions Biomediques August Pi Sunyer (IDIBAPS) and Hospital Clinic Barcelona, Barcelona, Spain., Martinez-Lapiscina EH; Institut d'Investigacions Biomediques August Pi Sunyer (IDIBAPS) and Hospital Clinic Barcelona, Barcelona, Spain., Llufriu S; Institut d'Investigacions Biomediques August Pi Sunyer (IDIBAPS) and Hospital Clinic Barcelona, Barcelona, Spain., Saiz A; Institut d'Investigacions Biomediques August Pi Sunyer (IDIBAPS) and Hospital Clinic Barcelona, Barcelona, Spain., Blanco Y; Institut d'Investigacions Biomediques August Pi Sunyer (IDIBAPS) and Hospital Clinic Barcelona, Barcelona, Spain., Martinez-Heras E; Institut d'Investigacions Biomediques August Pi Sunyer (IDIBAPS) and Hospital Clinic Barcelona, Barcelona, Spain., Solana E; Institut d'Investigacions Biomediques August Pi Sunyer (IDIBAPS) and Hospital Clinic Barcelona, Barcelona, Spain., Bäcker-Koduah P; Charité Universitaetsmedizin Berlin, Berlin, Germany., Behrens J; Charité Universitaetsmedizin Berlin, Berlin, Germany., Kuchling J; Charité Universitaetsmedizin Berlin, Berlin, Germany., Asseyer S; Charité Universitaetsmedizin Berlin, Berlin, Germany.; Max Delbrueck Center for Molecular Medicine, Berlin, Germany., Scheel M; Charité Universitaetsmedizin Berlin, Berlin, Germany., Chien C; Charité Universitaetsmedizin Berlin, Berlin, Germany.; Max Delbrueck Center for Molecular Medicine, Berlin, Germany., Zimmermann H; Charité Universitaetsmedizin Berlin, Berlin, Germany.; Max Delbrueck Center for Molecular Medicine, Berlin, Germany., Motamedi S; Charité Universitaetsmedizin Berlin, Berlin, Germany., Kauer-Bonin J; Charité Universitaetsmedizin Berlin, Berlin, Germany., Brandt A; Charité Universitaetsmedizin Berlin, Berlin, Germany., Saez-Rodriguez J; Institute for Computational Biomedicine, Heidelberg University Hospital, and Heidelberg University, Heidelberg, Germany., Alexopoulos LG; ProtATonce Ltd, Athens, Greece.; School of Mechanical Engineering, National Technical University of Athens, Zografou, Greece., Paul F; Charité Universitaetsmedizin Berlin, Berlin, Germany.; Max Delbrueck Center for Molecular Medicine, Berlin, Germany., Harbo HF; University of Oslo, Oslo, Norway.; Oslo University Hospital, Oslo, Norway., Shams H; Department of Neurology, University of California, San Francisco, USA., Oksenberg J; Department of Neurology, University of California, San Francisco, USA., Uccelli A; Department of Neurosciences, Rehabilitation, Ophthalmology, Genetics, Maternal and Child Health, University of Genoa, Genoa, Italy.; IRCCS Ospedale Policlinico San Martino, Genoa, Italy., Baeza-Yates R; School of Engineering, Pompeu Fabra University, Barcelona, Spain., Villoslada P; Department of Medicine and Life Sciences, Pompeu Fabra University, Barcelona, Spain. pablo.villoslada@upf.edu.; Hospital del Mar Research Institute, Barcelona, Spain. pablo.villoslada@upf.edu.
Jazyk: angličtina
Zdroj: Journal of neurology [J Neurol] 2024 Mar; Vol. 271 (3), pp. 1133-1149. Date of Electronic Publication: 2023 Dec 22.
DOI: 10.1007/s00415-023-12132-z
Abstrakt: Background: Multiple sclerosis patients would benefit from machine learning algorithms that integrates clinical, imaging and multimodal biomarkers to define the risk of disease activity.
Methods: We have analysed a prospective multi-centric cohort of 322 MS patients and 98 healthy controls from four MS centres, collecting disability scales at baseline and 2 years later. Imaging data included brain MRI and optical coherence tomography, and omics included genotyping, cytomics and phosphoproteomic data from peripheral blood mononuclear cells. Predictors of clinical outcomes were searched using Random Forest algorithms. Assessment of the algorithm performance was conducted in an independent prospective cohort of 271 MS patients from a single centre.
Results: We found algorithms for predicting confirmed disability accumulation for the different scales, no evidence of disease activity (NEDA), onset of immunotherapy and the escalation from low- to high-efficacy therapy with intermediate to high-accuracy. This accuracy was achieved for most of the predictors using clinical data alone or in combination with imaging data. Still, in some cases, the addition of omics data slightly increased algorithm performance. Accuracies were comparable in both cohorts.
Conclusion: Combining clinical, imaging and omics data with machine learning helps identify MS patients at risk of disability worsening.
(© 2023. The Author(s).)
Databáze: MEDLINE
Externí odkaz: