The integration of clinical data in the assessment of multiple sclerosis – A review. (June 2022)

• Record Type: Journal Article
• Authors: Ostellino, Sofia; Benso, Alfredo; Politano, Gianfranco
• Is Part Of: Computer methods and programs in biomedicine. Volume 221(2022)
• Publication Date: 2022-06
• Subjects: Multiple sclerosis -- Clinical data -- Computational integration -- Neuropsychological tests -- Cognitive assessment -- Personalized medicine -- Bio-markers; Medicine -- Computer programs -- Periodicals; Biology -- Computer programs -- Periodicals; Computers -- Periodicals; Medicine -- Periodicals; Médecine -- Logiciels -- Périodiques; Biologie -- Logiciels -- Périodiques; Biology -- Computer programs; Medicine -- Computer programs; Periodicals; Electronic journals; 610.28
• DOI: 10.1016/j.cmpb.2022.106900
• Languages: English
• ISSNs: 0169-2607

Abstract

Highlights: Multiple Sclerosis (MS) progression needs to be monitored through heterogeneous clinical measures. Cognitive assessment via neuropsychological tests (NP) is fundamental and informative in MS monitoring. Computerized NPs allow good quality and efficient cognitive examinations. Computational integration of data related to different neurodegenerative disease domains is a promising research field. Artificial Intelligence is, at the moment, the most promising computational approach to extract new knowledge from large sets of available clinical data. Abstract: Background and Objectives: Multiple Sclerosis (MS) is a neurological disease associated with various and heterogeneous clinical characteristics. Given its complex nature and its unpredictable evolution over time, there isn't an established and exhaustive clinical protocol (or tool) for its diagnosis nor for monitoring its progression. Instead, different clinical exams and physical/psychological evaluations need to be taken into account. The Expanded Disability Status Scale (EDSS) is the most used clinical scale, but it suffers from several limitations. Developing computational solutions for the identification of bio-markers of disease progression that overcome the downsides of currently used scales is crucial and is gaining interest in current literature and research. Methods: This Review focuses on the importance of approaching MS diagnosis and monitoring by investigating correlations between cognitive impairment and clinical data that refer to different MS domains. We review papers that integrate heterogeneous data and analyse them with statistical methods to understand their applicability into more advanced computational tools. Particular attention is paid to the impact that computational approaches can have on personalized-medicine. Results: Personalized medicine for neuro-degenerative diseases is an unmet clinical need which can be addressed using computational approaches able to efficiently integrate heterogeneous clinical data extracted from both private and publicly available electronic health databases. Conclusions: Reliable and explainable Artificial Intelligence are computational approaches required to understand the complex and demonstrated interactions between MS manifestations as well as to provide reliable predictions on the disease evolution, representing a promising research field.