Classifying multiple sclerosis patients on the basis of SDMT performance using machine learning

2020 ◽  
Vol 27 (1) ◽  
pp. 107-116
Author(s):  
Korhan Buyukturkoglu ◽  
Dana Zeng ◽  
Srinidhi Bharadwaj ◽  
Ceren Tozlu ◽  
Enricomaria Mormina ◽  
...  
Keyword(s):  
Multiple Sclerosis ◽  
Functional Connectivity ◽  
Resting State ◽  
Diffusion Tensor ◽  
Mean Diffusivity ◽  
Cognitive Status ◽  
Lesion Volume ◽  
Cognitive Changes ◽  
Final Model ◽  
Multiple Sclerosis Patients

Objective: To build a model to predict cognitive status reflecting structural, functional, and white matter integrity changes in early multiple sclerosis (MS). Methods: Based on Symbol Digit Modalities Test (SDMT) performance, 183 early MS patients were assigned “lower” or “higher” performance groups. Three-dimensional (3D)-T2, T1, diffusion weighted, and resting-state magnetic resonance imaging (MRI) data were acquired in 3T. Using Random Forest, five models were trained to classify patients into two groups based on 1—demographic/clinical, 2—lesion volume/location, 3—local/global tissue volume, 4—local/global diffusion tensor imaging, and 5—whole-brain resting-state-functional-connectivity measures. In a final model, all important features from previous models were concatenated. Area under the receiver operating characteristic curve (AUC) values were calculated to evaluate classifier performance. Results: The highest AUC value (0.90) was achieved by concatenating all important features from neuroimaging models. The top 10 contributing variables included volumes of bilateral nucleus accumbens and right thalamus, mean diffusivity of left cingulum-angular bundle, and functional connectivity among hubs of seven large-scale networks. Conclusion: These results provide an indication of a non-random brain pattern mostly compromising areas involved in attentional processes specific to patients who perform worse in SDMT. High accuracy of the final model supports this pattern as a potential neuroimaging biomarker of subtle cognitive changes in early MS.

Diffusion tensor magnetic resonance imaging in very early onset pediatric multiple sclerosis

2015 ◽  
Vol 22 (5) ◽  
pp. 620-627 ◽  
Author(s):  
MA Rocca ◽  
M Sonkin ◽  
M Copetti ◽  
E Pagani ◽  
DL Arnold ◽  
...  
Keyword(s):  
Magnetic Resonance Imaging ◽  
Multiple Sclerosis ◽  
Magnetic Resonance ◽  
Structural Integrity ◽  
Linear Models ◽  
Diffusion Tensor ◽  
Mean Diffusivity ◽  
Lesion Volume ◽  
Resonance Imaging ◽  
The Impact

Objectives: Active myelination during childhood may influence the impact of multiple sclerosis (MS) on brain structural integrity. We studied normal-appearing white matter (NAWM) in children with MS onset before age 12 years using diffusion tensor (DT) magnetic resonance imaging (MRI). Methods: DT MRI scans were obtained from 22 MS children with their first attack before age 12 years, and 31 healthy controls from two referral centers. Using probabilistic tractography, brain tissue integrity within interhemispheric, intrahemispheric, and projection tracts was compared between patients and site-matched controls. The impact of disease and age at MRI on tract NAWM fractional anisotropy (FA) and mean diffusivity (MD) values was evaluated using linear models. Results: Compared to controls, pediatric MS patients had reduced FA and increased MD of the bilateral superior longitudinal fasciculus and corpus callosum (CC), without center-by-group interaction. CC NAWM average FA was correlated with brain T2 lesion volume. In controls, the majority of the tracts analyzed showed a significant increase of FA and decrease of MD with age. Such a linear correlation was lost in patients. Conclusions: In very young pediatric MS patients, DT MRI abnormalities affect brain WM tracts differentially, and are only partially correlated with focal WM lesions. Impaired maturation of WM tracts with age may be an additional factor contributing to these findings.

scholarly journals Effect of Maternal Supplementation of Lutein on Brain Functional Organization of Offspring in a Piglet Model

2020 ◽  
Vol 4 (Supplement_2) ◽  
pp. 105-105
Author(s):  
Xi Fang ◽  
Wenwu Sun ◽  
Julie Jeon ◽  
Srujana Rayalam ◽  
Qun Zhao ◽  
...  
Keyword(s):  
Cognitive Development ◽  
Functional Connectivity ◽  
Resting State ◽  
Functional Organization ◽  
Diffusion Tensor ◽  
Control Diet ◽  
Mean Diffusivity ◽  
Superior Temporal Gyrus ◽  
Late Gestation ◽  
Maternal Supplementation

Abstract Objectives Lutein preferentially accumulates in human eyes and brains across the lifespan and is associated with visual and cognitive function. Dietary lutein intake during gestation and lactation may influence the development of neuronal networks of the infants. This study aims to provide preliminary data on the effect of maternal lutein supplementation during perinatal period on brain functional organization of the offspring. Methods Pregnant sows (n = 6) were fed a corn-based control diet (CON) or CON supplemented with lutein (LUT, 2 mg/kg BW/day) from late gestation to lactation for 60 days. Piglets (n = 7 in each group) underwent magnetic resonance imaging (MRI) to acquire anatomical, diffusion tensor imaging, and resting-state functional MRI (rs-fMRI) data at weaning (21d old). Using a sparse dictionary learning approach, six resting-state networks were examined that resembles that of humans. Results Piglets from LUT-fed sows showed a 7.7% decreased functional connectivity in executive control network and 13.2% decrease in cerebellum network compared to that of CON piglets, suggesting perinatal LUT supplementation may suppress the activation of executive and motor function at resting state in infants. In addition, piglets from LUT-fed sows had a 10.3% decrease in functional connectivity in the visual network encompassing primary, secondary, and associative visual cortex and an 8% lower connectivity within auditory network encompassing superior temporal gyrus and auditory cortex in comparison to that of CON piglets, indicating maternal LUT supplementation may lead to a lower functional activation of the intrinsic visual and auditory networks of offspring at weaning. Slight changes in sensorimotor and default mode networks were also observed. Moreover, perinatal lutein supplementation did not change mean diffusivity, fractional anisotropy, and fiber length in hippocampus, the key component of memory formation and cognitive development. Conclusions Maternal supplementation of lutein may alter the functional organization of the offspring within multiple intrinsic networks at resting state that may underlie the functional outcomes of cognitive development of the offspring at weaning. Funding Sources Georgia Experimental Agricultural Station, Faculty research grant from Office or Research at the UGA, and Division of Research at PCOM.

scholarly journals The link between resting-state functional connectivity and cognition in MS patients

2013 ◽  
Vol 20 (3) ◽  
pp. 338-348 ◽  
Author(s):  
Álvaro J Cruz-Gómez ◽  
Noelia Ventura-Campos ◽  
Antonio Belenguer ◽  
Cesar Ávila ◽  
Cristina Forn
Keyword(s):  
Functional Connectivity ◽  
Resting State ◽  
Cognitive Status ◽  
Control Group ◽  
Cognitively Impaired ◽  
Lesion Volume ◽  
Resting State Functional Connectivity ◽  
Frontoparietal Network ◽  
Brain Parenchymal Fraction ◽  
Brain Parenchymal

Objective: The objective of this paper is to explore differences in resting-state functional connectivity between cognitively impaired and preserved multiple sclerosis (MS) patients. Methods: Sixty MS patients and 18 controls were assessed with the Brief Repeatable Battery of Neuropsychological Tests (BRB-N). A global Z score of the BRB-N was obtained and allowed us to classify MS patients as cognitively impaired and cognitively preserved ( n = 30 per group). Functional connectivity was assessed by independent component analysis of resting-state networks (RSNs) related to cognition: the default mode network, left and right frontoparietal and salience network. Between-group differences were evaluated and a regression analysis was performed to describe relationships among cognitive status, functional connectivity and radiological variables. Results: Compared to cognitively preserved patients and healthy controls, cognitively impaired patients showed a lesser degree of functional connectivity in all RSNs explored. Cognitively preserved patients presented less connectivity than the control group in the left frontoparietal network. Global Z scores were positively and negatively correlated with brain parenchymal fraction and lesion volume, respectively. Conclusion: Decreased cognitive performance is accompanied by reduced resting state functional connectivity and directly related to brain damage. These results support the use of connectivity as a powerful tool to monitor and predict cognitive impairment in MS patients.

Distributed changes in default-mode resting-state connectivity in multiple sclerosis

2011 ◽  
Vol 17 (4) ◽  
pp. 411-422 ◽  
Author(s):  
Simona Bonavita ◽  
Antonio Gallo ◽  
Rosaria Sacco ◽  
Marida Della Corte ◽  
Alvino Bisecco ◽  
...  
Keyword(s):  
Multiple Sclerosis ◽  
Resting State ◽  
Cingulate Cortex ◽  
Posterior Cingulate Cortex ◽  
Cognitive Status ◽  
Anterior Cingulate ◽  
Default Mode ◽  
Relapsing Remitting ◽  
Multiple Sclerosis Patients ◽  
Global And Local

Background: The default-mode network (DMN) has been increasingly recognized as relevant to cognitive status. Objectives: To explore DMN changes in patients with relapsing–remitting (RR) multiple sclerosis (MS) and to relate these to the cognitive status. Methods: Eighteen cognitively impaired (CI) and eighteen cognitively preserved (CP) RRMS patients and eighteen healthy controls (HCs), matched for age, sex and education, underwent neuropsychological evaluation and anatomical and resting-state functional MRI (rs-fMRI). DMN functional connectivity was evaluated from rs-fMRI data via independent component analysis. T2 lesion load (LL) was computed by a semi-automatic method and global and local atrophy was estimated by SIENAX and SPM8 voxel-based morphometry analyses from 3D-T1 images. Results: When the whole group of RRMS patients was compared with HCs, DMN connectivity was significantly weaker in the anterior cingulate cortex, whereas it was significantly weaker in the core but stronger at the periphery of the posterior cingulate cortex. These findings were more evident in CP than CI patients. Observed DMN changes did not correlate with global atrophy or T2-LL, but were locally associated with regional grey matter loss. Conclusion: Relapsing–remitting multiple sclerosis patients show a consistent dysfunction of DMN at the level of the anterior node. DMN distribution changes in the posterior node may reflect a possible compensatory effect on cognitive performance.

Expansion of chronic lesions is linked to disease progression in relapsing–remitting multiple sclerosis patients

2020 ◽  
pp. 135245852097435
Author(s):  
Samuel Klistorner ◽  
Michael H Barnett ◽  
Con Yiannikas ◽  
Joshua Barton ◽  
John Parratt ◽  
...  
Keyword(s):  
Multiple Sclerosis ◽  
Disease Progression ◽  
Diffusion Tensor ◽  
Brain Lesion ◽  
Lesion Volume ◽  
Free Standing ◽  
Relapsing Remitting ◽  
Remitting Multiple Sclerosis ◽  
Multiple Sclerosis Patients ◽  
Relapsing Remitting Multiple Sclerosis

Background: Slow-burning inflammation is putatively associated with lesion expansion and leads to progressive loss of axons and disability worsening. Objective: To investigate the incidence and extent of chronic white matter lesion expansion in relapsing–remitting multiple sclerosis (RRMS) patients and to evaluate its relationship with biomarkers of disease progression. Methods: Pre- and post-gadolinium T1, fluid-attenuated inversion recovery (FLAIR) and diffusion tensor images were acquired from 33 patients. Lesional activity were analysed between baseline and 48 months using custom-designed software. Results: A total of 569 lesions were identified as chronic at baseline, of which 261 were expanding, 236 were stable and 72 were shrinking. In addition, 139 new lesions (both confluent and free-standing) were observed. Chronic lesion expansion was associated with patient’s age and accounted for the bulk (67.3%) of total brain lesion volume increase, while only 32.7% was attributable to new lesion formation. Change in chronic lesion volume correlated with the rate of brain atrophy ( r = −0.57, p = 0.001), change of Expanded Disability Status Scale (EDSS; r = 0.38, p = 0.03) and an increase of isotropic diffusivity inside the lesions ( r = 0.75, p < 0.001). Conclusion: Expansion of chronic lesions in RRMS patients is the primary determinant of increased T2 total lesion load. It significantly contributes to disease progression and partially driving axonal loss inside the lesions and brain damage outside of lesional tissue.

Changes of brain resting state functional connectivity predict the persistence of cognitive rehabilitation effects in patients with multiple sclerosis

2013 ◽  
Vol 20 (6) ◽  
pp. 686-694 ◽  
Author(s):  
Laura Parisi ◽  
Maria A Rocca ◽  
Flavia Mattioli ◽  
Massimiliano Copetti ◽  
Ruggero Capra ◽  
...  
Keyword(s):  
Multiple Sclerosis ◽  
Functional Connectivity ◽  
Resting State ◽  
Cognitive Rehabilitation ◽  
Severe Depression ◽  
Control Group ◽  
Neuropsychological Performance ◽  
Resting State Functional Connectivity ◽  
Multiple Sclerosis Patients

Objective: We investigated whether the efficacy of 12-week cognitive rehabilitation in MS patients persists six months after treatment termination and, together with resting state (RS) functional connectivity (FC), changes on neuropsychological performance at follow-up. Methods: Eighteen MS patients with cognitive deficits, assigned randomly either to undergo treatment ( n=9) or not ( n=9), underwent neuropsychological evaluation at baseline (t0), after 12 weeks of rehabilitation (t1) and at six-month follow-up (t2). RS fMRI was obtained at t0 and t1. Changes in neuropsychological performance and their correlations with RS FC modifications were assessed using longitudinal linear models. Results: At t2 vs. t0, compared with the control group, treated group patients improved in tests of attention, executive function, depression and quality of life (QoL). Neuropsychological scores in these tests at t2 were significantly correlated with RS FC changes in cognitive-related networks and RS FC of the anterior cingulum. RS FC changes in the default mode network predicted cognitive performance and less severe depression, whereas RS FC changes of the executive network predicted better QoL. Discussion: Changes in RS FC of cognitive-related networks helps to explain the persistence of the effects of cognitive rehabilitation after several months in relapsing–remitting multiple sclerosis patients and their improvement on depression and QoL scales.

scholarly journals Relationship between timed 25-foot walk and diffusion tensor imaging in multiple sclerosis

2016 ◽  
Vol 2 ◽  
pp. 205521731665536 ◽  
Author(s):  
Sylvia Klineova ◽  
Rebecca Farber ◽  
Catarina Saiote ◽  
Colleen Farrell ◽  
Bradley N Delman ◽  
...  
Keyword(s):  
Magnetic Resonance Imaging ◽  
Multiple Sclerosis ◽  
Diffusion Tensor Imaging ◽  
Magnetic Resonance ◽  
Diffusion Tensor ◽  
Mean Diffusivity ◽  
Resonance Imaging ◽  
Corticospinal Tracts ◽  
Multiple Sclerosis Patients ◽  
And Diffusion

Objective/Background The majority of multiple sclerosis patients experience impaired walking ability, which impacts quality of life. Timed 25-foot walk is commonly used to gauge gait impairment but results can be broadly variable. Objective biological markers that correlate closely with patients’ disability are needed. Diffusion tensor imaging, quantifying fiber tract integrity, might provide such information. In this project we analyzed relationships between timed 25-foot walk, conventional and diffusion tensor imaging magnetic resonance imaging markers. Design/Methods A cohort of gait impaired multiple sclerosis patients underwent brain and cervical spinal cord magnetic resonance imaging. Diffusion tensor imaging mean diffusivity and fractional anisotropy were measured on the brain corticospinal tracts and spinal restricted field of vision at C2/3. We analyzed relationships between baseline timed 25-foot walk, conventional and diffusion tensor imaging magnetic resonance imaging markers. Results Multivariate linear regression analysis showed a statistically significant association between several magnetic resonance imaging and diffusion tensor imaging metrics and timed 25-foot walk: brain mean diffusivity corticospinal tracts (p = 0.004), brain corticospinal tracts axial and radial diffusivity (P = 0.004 and 0.02), grey matter volume (p = 0.05), white matter volume (p = 0.03) and normalized brain volume (P = 0.01). The linear regression model containing mean diffusivity corticospinal tracts and controlled for gait assistance was the best fit model (p = 0.004). Conclusions Our results suggest an association between diffusion tensor imaging metrics and gait impairment, evidenced by brain mean diffusivity corticospinal tracts and timed 25-foot walk.

scholarly journals Classifying minimally disabled multiple sclerosis patients from resting state functional connectivity

NeuroImage ◽  
2012 ◽  
Vol 62 (3) ◽  
pp. 2021-2033 ◽  
Author(s):  
Jonas Richiardi ◽  
Markus Gschwind ◽  
Samanta Simioni ◽  
Jean-Marie Annoni ◽  
Beatrice Greco ◽  
...  
Keyword(s):  
Multiple Sclerosis ◽  
Functional Connectivity ◽  
Resting State ◽  
Resting State Functional Connectivity ◽  
Multiple Sclerosis Patients

scholarly journals Concurrent anatomical, physiological and network changes in cognitively impaired multiple sclerosis patients

2020 ◽  
Author(s):  
Danka Jandric ◽  
Ilona Lipp ◽  
David Paling ◽  
David Rog ◽  
Gloria Castellazzi ◽  
...  
Keyword(s):  
Multiple Sclerosis ◽  
Blood Flow ◽  
Cerebral Blood Flow ◽  
Functional Connectivity ◽  
Resting State ◽  
Metabolic Stress ◽  
Cognitively Impaired ◽  
Anatomical Connectivity ◽  
Resting State Network ◽  
Multiple Sclerosis Patients

AbstractCognitive impairment in multiple sclerosis is associated with functional connectivity abnormalities, but the pathological substrates of these abnormalities are not well understood. It has been proposed that resting-state network nodes that integrate information from disparate regions are susceptible to metabolic stress, which may impact functional connectivity. In multiple sclerosis, pathology could increase metabolic stress within axons, damaging the anatomical connections of network regions, and leading to functional connectivity changes. We tested this hypothesis by assessing whether resting state network regions that show functional connectivity abnormalities in people with cognitive impairment also show anatomical connectivity abnormalities.Multimodal MRI and neuropsychological assessments were performed in 102 relapsing remitting multiple sclerosis patients and 27 healthy controls. Patients were considered cognitively impaired if they obtained a z-score of ≤1.5 on ≥2 tests of the Brief Repeatable Battery of Neuropsychological Tests (n=55). Functional connectivity was assessed with Independent Component Analysis of resting state fMRI images, and anatomical connectivity with Anatomical Connectivity Mapping of diffusion-weighted MRI. Exploratory analyses of fractional anisotropy and cerebral blood flow changes were conducted to assess local tissue characteristics.We found significantly decreased functional connectivity in the anterior and posterior default mode networks and significant increases in the right and left frontoparietal networks in cognitively impaired relative to cognitively preserved patients. Networks showing functional abnormalities also showed reduced anatomical connectivity and white matter microstructure integrity as well as reduced local tissue cerebral blood flow.Our results identify key pathological correlates of functional connectivity abnormalities associated with impaired cognitive function in multiple sclerosis, consistent with metabolic dysfunction in functional network regions.

Clinical significance of atrophy and white matter mean diffusivity within the thalamus of multiple sclerosis patients

2013 ◽  
Vol 19 (11) ◽  
pp. 1478-1484 ◽  
Author(s):  
Ralph HB Benedict ◽  
Hanneke E Hulst ◽  
Niels Bergsland ◽  
Menno M Schoonheim ◽  
Michael G Dwyer ◽  
...  
Keyword(s):  
Multiple Sclerosis ◽  
Cognitive Impairment ◽  
White Matter ◽  
Cognitive Processing ◽  
Diffusion Tensor ◽  
Mean Diffusivity ◽  
Brain Magnetic Resonance Imaging ◽  
Cognitive Tests ◽  
Explained Variance ◽  
Multiple Sclerosis Patients

Background: Gray-matter (GM) atrophy is strongly predictive of cognitive impairment in multiple sclerosis (MS) patients. The thalamus is the region where the atrophy/cognition correlation is most robust. However, few studies have assessed diffusion tensor imaging (DTI) metrics within the thalamus. Objective: This study was designed to determine if thalamus white matter DTI predicts cognitive impairment after accounting for the effects of volume loss. Methods: We enrolled 75 MS patients and 18 healthy controls undergoing 3T brain magnetic resonance imaging (MRI). Thalamus volumes were calculated on 3D T1 images. Voxelwise analyses of DTI metrics were performed within the thalamic white matter tracts. Neuropsychological (NP) testing, acquired using consensus standard methods, contributed measures of memory, cognitive processing speed and executive function. Results: All cognitive tests were significantly predicted ( R2 =0.31, p<0.001) by thalamus volume after accounting for influence of demographics. Mean diffusivity was retained in regression models predicting all cognitive tests, adding from 7–13% of additional explained variance ( p<0.02) after accounting for thalamus volume. Conclusions: We confirm the significant role of thalamus atrophy in MS-associated cognitive disorder, and further report that subtle thalamus pathology as detected by DTI adds incremental explained variance in predicting cognitive impairment.

Sign in / Sign up

Export Citation Format

Share Document