Three cognitive phenotypes with differential patterns of brain atrophy measures integrate neuropsychological testing with MRI measures into a clinically applicable framework that may help bridge the gap between structural imaging findings and everyday cognitive assessment in PwMS.
Abstract
Background Cognitive impairment (CI) is common in multiple sclerosis (MS) yet poorly captured by conventional disability scales. Although neuropsychological assessment and magnetic resonance imaging (MRI) are routinely used separately, there is no simple clinically applicable framework integrating cognitive performance with structural brain changes to identify patients at increased risk of cognitive decline. Integrating neuropsychological testing with MRI-based atrophy metrics may yield clinically useful cognitive phenotypes with differential patterns of brain atrophy measures. Methods Data were collected from 79 patients with multiple sclerosis (PwMS) who underwent comprehensive neuropsychological assessment and brain MRI. Neuropsychological variables were subjected to a feature selection procedure based on variance and quartile coefficient of dispersion filtering, followed by Pearson correlation and mutual information (MI) analyses to generate reduced feature sets. These feature sets were used as input for unsupervised clustering with the Partitioning Around Medoids (PAM) algorithm to identify cognitive phenotypes. Differences between the resulting groups in the degree of brain atrophy measures were subsequently evaluated using appropriate statistical tests—one-way ANOVA or the Kruskal–Wallis test. Post hoc analysis was performed using a pairwise t-test, Welch's t-test, or Wilcoxon test with the Holm-Bonferroni correction, depending on the data distribution and variance. Results The feature selection procedure based on variance and mutual information identified neuropsychological features that were subsequently used for clustering. Based on these features, the PAM algorithm identified three distinct groups of PwMS that differed in their clinical characteristics, degree of brain atrophy measures, and cognitive phenotype, ranging from preserved cognition to global cognitive impairment. Conclusion Three cognitive phenotypes with differential patterns of brain atrophy measures integrate neuropsychological testing with MRI measures into a clinically applicable framework that may help bridge the gap between structural imaging findings and everyday cognitive assessment in PwMS. This approach may improve screening, enable earlier detection of CI, improve monitoring, and provide valuable information for rehabilitation planning.
Multiple Sclerosis (MS) is a chronic neurological disorder of the central nervous system. Its diagnosis and monitoring rely heavily on Magnetic Resonance Imaging (MRI), while increasingly complemented by other imaging, signal-based, gait-related, and clinical data sources. Manual interpretation is labor-intensive and variable, driving interest in machine learning (ML) approaches for automated lesion segmentation, classification, and detection across MRI and complementary modalities. This systematic review, conducted under PRISMA guidelines, consolidates recent advances in ML for MS diagnosis and monitoring and proposes a task-based taxonomy linking clinical objectives to methodological families. The analysis reveals that most studies rely solely on structural MRI, with limited exploration of multimodal or longitudinal data, including combinations with optical coherence tomography, retinal imaging, electroencephalography, gait analysis, plantar pressure, and clinical variables. The Dice Similarity Coefficient (DSC) dominates evaluation, with reported scores reaching up to 0.981. Results are benchmarked on public datasets such as ISBI 2015 and MICCAI 2016, and the review critically examines reproducibility barriers posed by small or private cohorts. Despite notable progress, key bottlenecks—dataset scarcity, class imbalance, domain shift, and inconsistent evaluation—continue to hinder generalization and clinical adoption. Computational costs and the absence of standardized pipelines further delay translation. Emerging strategies are also highlighted, including federated learning for privacy-preserving training, few-shot learning for data-efficient modeling, multimodal integration, and explainable AI to support transparency and clinician trust in comprehensive MS analysis. By unifying methodological insights into a structured taxonomy, benchmarking progress across datasets, and charting pathways to clinical deployment, this review provides a comprehensive roadmap toward building robust, data-efficient, and trustworthy ML systems for clinical decision support in MS.
Marthe Elgawly, Beyza Nur Elaslan, Marco Cascio et al.· IEEE Access· 0 citations
Cognitive impairment and dementia represent a heterogeneous group of disorders with variable patterns of brain atrophy. Automated brain volumetry combined with machine learning offers the potential to identify biologically meaningful subgroups beyond traditional clinical classifications. This study aimed to characterize cerebral volumetric profiles and identify distinct volumetric endophenotypes in individuals with cognitive impairment using unsupervised machine learning. This retrospective cross-sectional study included 78 participants with cognitive impairment who underwent T1-weighted three-dimensional magnetic resonance imaging (MRI). Automated brain volumetry was performed using mdbrain software (version 2.2), which provides normalized percentile values for 42 brain regions adjusted for age, sex, and total intracranial volume. Unsupervised k-means clustering was applied to volumetric percentiles to identify distinct endophenotypes. Cluster validation was performed using elbow method, silhouette analysis, Calinski-Harabasz index, and Davies-Bouldin index. Discriminative features were identified using ANOVA with effect size calculations (η²). A multinomial logistic regression model was developed to predict cluster membership. The cohort comprised 48 (61.5%) females and 30 (38.5%) males. Substantial inter-individual variability was observed across all brain regions, with hippocampal percentiles ranging from 0.1 to 99.8 (median: 25.6). Clustering analysis revealed three distinct volumetric endophenotypes: Cluster 1 (Preserved, n=27, 34.6%) characterized by percentiles above the 40th percentile; Cluster 2 (Moderate Atrophy, n=31, 39.7%) with percentiles in the 10th–30th range; and Cluster 3 (Severe Atrophy, n=20, 25.6%) with hippocampal percentiles below the 5th percentile and ventricular percentiles above the 85th percentile. Hippocampal volumes demonstrated the strongest discriminative ability (left hippocampus: F=68.4, η²=0.64; right hippocampus: F=65.2, η²=0.63), followed by temporal (η²=0.60–0.61) and ventricular volumes (η²=0.57–0.58). Interhemispheric asymmetry was minimal (median absolute difference: 3.4–7.2 percentile points). A parsimonious model using four volumetric features (left hippocampus, left temporal, left ventricle, and right frontal percentiles) achieved excellent classification accuracy (87.2%, cross-validated AUC=0.91). Automated brain volumetry combined with unsupervised machine learning identified three distinct volumetric endophenotypes in cognitive impairment, primarily discriminated by hippocampal, temporal, and ventricular volumes. These findings suggest that cognitive impairment manifests through heterogeneous patterns of structural brain involvement, with potential implications for prognosis and personalized care. The high accuracy of a parsimonious four-feature model supports the clinical utility of focused volumetric assessment for risk stratification. Future longitudinal studies with multimodal biomarker integration are warranted to validate these findings and establish their prognostic value.
Albert Vamanu, L. Borza, Livia Livinț-Popa et al.· Balneo and PRM Research Jour...· 0 citations
Background/Objectives: This study evaluated automated magnetic resonance imaging (MRI) volumetry for characterizing structural brain changes in Alzheimer’s disease (AD), mild cognitive impairment (MCI), and cognitively healthy controls (HC), and examined its association with cognitive performance. Methods: This retrospective observational study included consecutively enrolled individuals aged ≥65 years. AD dementia and MCI were diagnosed according to the 2011 National Institute on Aging–Alzheimer’s Association (NIA-AA) clinical criteria. Automated volumetric analysis of structural MRI was used to obtain the total intracranial volume-normalized cortical and subcortical volumes, cerebrospinal fluid (CSF) compartments, and hemispheric asymmetry indices. Between-group comparisons were corrected using the Benjamini–Hochberg false discovery rate. Prespecified volumetric markers were evaluated using receiver operating characteristic analysis and internally validated logistic regression models. Results: The study included 102 participants (34 per group). Compared with HC, the AD group showed lower total cerebrum, right hippocampal, and anterior cingulate gyrus (ACgG) volumes and higher CSF volumes. Compared with MCI, the AD group exhibited lower thalamic and caudate volumes. CSF volume showed the highest individual discriminative performance for differentiating AD from HC (AUC = 0.837), whereas the combined hippocampal–ACgG–CSF model showed the best performance for differentiating MCI from HC (AUC = 0.826). After adjustment for diagnostic group, cognitive performance remained positively associated with hippocampal volume and negatively with CSF and lateral ventricular volumes. Conclusions: Automated MRI volumetry may support the quantitative assessment of neurodegeneration in clinically defined AD and MCI. Combined volumetric markers improved discrimination between MCI and HC, although these findings require external validation in larger, longitudinal, biomarker-characterized cohorts.
A. Alagoz, Serap Ozturk, S. D. Bunul et al.· Diagnostics· 0 citations
Cognitive decline associated with Alzheimer's disease (AD) and mild cognitive impairment (MCI) represents one of the most critical public health challenges of the twenty-first century, affecting tens of millions of people worldwide. This paper presents a comparative study of machine learning algorithms applied to early detection of cognitive decline using the open access series of imaging studies (OASIS-3) dataset, a longitudinal multimodal neuroimaging and clinical cohort of 1378 participants aged 42-95 years. To avoid data leakage, all features directly encoding a clinical diagnosis, including clinical dementia rating (CDR) scores, diagnostic (DX) codes, and mini-mental state examination (MMSE) results, are excluded; instead, fifteen structural neuroimaging features describing brain morphology (e.g., hippocampal and entorhinal measurements) and neuropsychological sub-test features are selected via a two-stage recursive feature elimination and greedy forward-search procedure. We evaluate six supervised models: logistic regression, support vector machine (SVM), random forest, gradient boosting, K-nearest neighbours (KNN), and CatBoost. Gradient boosting achieves the highest weighted F1-score (0.856) on the held-out test set, followed by CatBoost (0.8482); CatBoost offers approximately $2\times$ lower inference latency. A neurosymbolic post-processing stage using the Scallop probabilistic logic engine identifies 166 participants with measurable structural brain atrophy (reduction in hippocampal volume) prior to a formal clinical diagnosis, demonstrating the added value of logic-augmented classifiers for early risk stratification. Feature-importance analysis identifies white-matter hyperintensity volume (MRI-visible lesions associated with vascular and neurodegenerative damage) and entorhinal/hippocampal measurements as the dominant predictors, consistent with established neurobiological evidence.
Sofiia A. Mikhailova, O. A. Mikhailova· KyivAcademUs2026· 0 citations
Comparison of different MRI modalities in predicting brain age and the relative importance of specific brain regions provides valuable insight into their relative predictive utility and their potential role in characterizing brain aging.
Hami Mahdavinataj, H. Sajedi, Seyed Amir Hossein Batouli· GeroScience· 0 citations
Summary Background Multiple sclerosis (MS) is increasingly recognised as a disorder of large-scale brain network reorganisation rather than a disease explained solely by focal demyelinating lesions. However, the relevance of structural network abnormalities to clinical heterogeneity and progression remains unclear. Methods We analysed 3T magnetic resonance imaging (MRI) from two independent MS cohorts (total n = 635): Dataset 1 from the Chinese neuroimmunological diseases (NIDBase) cohort (163 MS and 248 healthy controls [HC]) and Dataset 2 from the UK Biobank (117 MS and 107 HC). A subset of Dataset 1 underwent 256-channel resting-state high-density electroencephalography (hd-EEG) (135 MS and 80 HC). We constructed structural covariance networks (iSCNs) from 3D T1-weighted MRI using Kullback–Leibler similarity across 170 Automated Anatomical Labelling atlas 3 (AAL3) regions. Patients were stratified by disability, cognition, and disease activity; a subgroup with no evidence of disease activity (NEDA) with 1-year follow-up MRI (n = 33) was analysed longitudinally. Linear support vector machine classifiers evaluated topological and connectivity features for diagnosis and clinical stratification. Findings MS showed reproducible topological and connectivity abnormalities, involving thalamic and subcortical hubs and altered visual-network-related couplings. Disability showed the broadest abnormalities, whereas cognitive impairment and disease activity were associated with more selective changes. Patients meeting NEDA criteria showed subtle longitudinal nodal changes. Connectivity features performed best for MS-vs-HC discrimination, whereas topological features performed better for clinical stratification. Interpretation The iSCN approach identified reliable patterns of topological and connectivity impairment across MS and its clinical stratifications, providing insight into MS neuropathology and guiding future diagnostic and therapeutic biomarker development. Funding This work was funded by Beijing Research Ward Excellence Program (BRWEP2024W022010104, BRWEP2024W022010109), Beijing Scholar program (No. 106), the Project for Innovation and Development of Beijing Municipal Geriatric Medical Research Center (11000023T000002041657), Dengfeng Talent Program (DFL20220701), 10.13039/501100001809National Natural Science Foundation of China (82571539, 82501612), Xuanwu Hospital Talent Convergence Program–Leading Talents (HZ2021ZCLJ008), the Beijing Hospitals Authority's Ascent Plan (DFL20240801), and Beijing "Huizhi" Talent Program, Cultivation Program–Leading Talents (HZ2025PYLJ003).
