Machine Learning for Predicting Individual Severity of Blepharospasm Using Diffusion Tensor Imaging

Accumulating diffusion tensor imaging (DTI) evidence suggests that white matter abnormalities evaluated by local diffusion homogeneity (LDH) or fractional anisotropy (FA) occur in patients with blepharospasm (BSP), both of which are significantly correlated with disease severity. However, whether the individual severity of BSP can be identified using these DTI metrics remains unknown. We aimed to investigate whether a combination of machine learning techniques and LDH or FA can accurately identify the individual severity of BSP. Forty-one patients with BSP were assessed using the Jankovic Rating Scale and DTI. The patients were assigned to non-functionally and functionally limited groups according to their Jankovic Rating Scale scores. A machine learning scheme consisting of beam search and support vector machines was designed to identify non-functionally versus functionally limited outcomes, with the input features being LDH or FA in 68 white matter regions. The proposed machine learning scheme with LDH or FA yielded an overall accuracy of 88.67 versus 85.19% in identifying non-functionally limited versus functionally limited outcomes. The scheme also identified a sensitivity of 91.40 versus 85.87% in correctly identifying functionally limited outcomes, a specificity of 83.33 versus 83.67% in accurately identifying non-functionally limited outcomes, and an area under the curve of 93.7 versus 91.3%. These findings suggest that a combination of LDH or FA measurements and a sophisticated machine learning scheme can accurately and reliably identify the individual disease severity in patients with BSP.

Download Full-text

Machine Learning Classification of Traumatic Brain Injury Patients and Healthy Controls Using Multiple Indices of Diffusion Tensor Imaging

10.21203/rs.3.rs-42800/v1 ◽

2020 ◽

Author(s):

Hiba Abuelgasim Fadlelmoula Abdelrahman ◽

Shiho Ubukata ◽

Keita Ueda ◽

Gaku Fujimoto ◽

Naoya Oishi ◽

...

Keyword(s):

Machine Learning ◽

Traumatic Brain Injury ◽

Diffusion Tensor Imaging ◽

Brain Injury ◽

White Matter ◽

Diffusion Tensor ◽

Area Under The Curve ◽

Support Vector ◽

Healthy Controls

Abstract Background: Diffusion tensor imaging (DTI) indices provide quantitative measures of white matter microstructural changes following traumatic brain injury (TBI). However, there is still insufficient evidence for their use as predictive measures. Recently, there has been growing interest in using machine learning (ML) approaches to aid the diagnosis of many neurological and psychiatric illnesses including TBI. The aim of this study is to examine the potential of using multiple DTI indices in conjunction with ML to automate the classification of healthy subjects and patients with TBI across a spectrum of TBI severity.Methods: Participants were adult patients with chronic TBI (n=26) and age and gender-matched healthy controls (n=26). DTI images were obtained from all the participants. Tract-based spatial statistics (TBSS) analysis was applied to the DTI images. Classification models were built using principle component analysis (PCA) and support vector machines (SVM). Receiver operator characteristic (ROC) curve analysis and area under the curve (AUC) were used to assess the classification performance of the different classifiers.Results: The whole-brain white matter TBSS analyses showed significantly decreased FA, as well as increased MD, AD, and RD in TBI patients compared with healthy controls (all p-value < 0.01). The PCA and SVM-based ML classification using combined DTI indices classified TBI patients and healthy controls with the accuracy of 90.5% with an area under the curve (AUC) of 93 +/- 0.09.Conclusion: This study demonstrates the potential of a joint DTI and ML approach for objective classification of TBI patients and healthy controls.

Download Full-text

Abstract 46: Leukoaraiosis Evolves Through Discrete Acute Ischemic Injuries: Weekly Diffusion Tensor Imaging and T2 Relaxometry at 16 Time Points

Stroke ◽

10.1161/str.45.suppl_1.46 ◽

2014 ◽

Vol 45 (suppl_1) ◽

Author(s):

John Conklin ◽

Frank L Silver ◽

David J Mikulis ◽

Daniel M Mandell

Keyword(s):

Diffusion Tensor Imaging ◽

White Matter ◽

Diffusion Tensor ◽

White Matter Lesions ◽

Striking Similarity ◽

Aging Brain ◽

Large Artery ◽

Neurology Clinic ◽

Mri Scans ◽

The Individual

INTRODUCTION: Leukoaraiosis, the presence of “incidental” white matter lesions in the aging brain, is increasingly recognized as a predictor for dementia, ischemic stroke, intracerebral hemorrhage and vascular death. The pathogenesis of leukoaraiosis remains controversial, with abnormalities of small arterioles hypothesized to play an important role. To investigate this hypothesis, we sought to characterize the temporal evolution of the individual lesions making up leukoaraiosis. HYPOTHESIS: Discrete occlusive events at the level of small arterioles play a key role in the pathogenesis of leukoaraiosis. METHODS: Participants were prospectively recruited through an outpatient neurology clinic (inclusion criteria: age > 60 years, Fazekas grade 3 leukoaraiosis burden; exclusion criteria: cortical infarct, cardioembolic disease, dissection, carotid stenosis > 50%). Subjects underwent an identical MRI protocol in each of 16 consecutive weeks, including diffusion tensor imaging (DTI) and multi-echo T2-weighted imaging. Parametric maps of the apparent diffusion coefficient (ADC), fractional anisotropy (FA) and T2 relaxation time were constructed and coregistered (Analysis of Functional NeuroImages, NIH; 3D Slicer, www.slicer.org; Matlab, The MathWorks). Images were reviewed for new diffusion restricting lesions, and such lesions were manually segmented. Plots of lesion ADC, FA and T2 were generated and temporally aligned to the onset of acute diffusion restriction. RESULTS: Five subjects (mean age 69 ± 8 years) met criteria and completed all 16 MRI scans. There were no lacunar or large artery infarcts during the study period. A total of 9 new diffusion restricting white matter lesions were identified (mean volume 0.06 ± 0.03 cc). Evolution of these lesions showed striking similarity to that of cerebral infarction, with acute reduction in ADC, followed by gradual rise in ADC and T2, and corresponding decline in FA. At 8 weeks, new lesions were indistinguishable from pre-existing white matter disease. CONCLUSION: Leukoaraiosis evolves through temporally and spatially discrete acute ischemic injuries. This supports the hypothesized role of small vessel arteriolar pathology as a key pathogenetic mechanism.

Download Full-text

Prediction of the Clinical Severity of Progressive Supranuclear Palsy by Diffusion Tensor Imaging

Journal of Clinical Medicine ◽

10.3390/jcm9010040 ◽

2019 ◽

Vol 9 (1) ◽

pp. 40 ◽

Cited By ~ 1

Author(s):

Yao-Liang Chen ◽

Xiang-An Zhao ◽

Shu-Hang Ng ◽

Chin-Song Lu ◽

Yu-Chun Lin ◽

...

Keyword(s):

Machine Learning ◽

Diffusion Tensor Imaging ◽

Progressive Supranuclear Palsy ◽

Rating Scale ◽

Diffusion Tensor ◽

Learning Algorithm ◽

Mean Diffusivity ◽

Feature Reduction ◽

Clinical Severity ◽

Machine Learning Algorithm

Progressive supranuclear palsy (PSP) is characterized by a rapid and progressive clinical course. A timely and objective image-based evaluation of disease severity before standard clinical assessments might increase the diagnostic confidence of the neurologist. We sought to investigate whether features from diffusion tensor imaging of the entire brain with a machine learning algorithm, rather than a few pathogenically involved regions, may predict the clinical severity of PSP. Fifty-three patients who met the diagnostic criteria for probable PSP were subjected to diffusion tensor imaging. Of them, 15 underwent follow-up imaging. Clinical severity was assessed by the neurological examinations. Mean diffusivity and fractional anisotropy maps were spatially co-registered, normalized, and parcellated into 246 brain regions from the human Brainnetome atlas. The predictors of clinical severity from a stepwise linear regression model were determined after feature reduction by the least absolute shrinkage and selection operator. Performance estimates were obtained using bootstrapping, cross-validation, and through application of the model in the patients who underwent repeated imaging. The algorithm confidently predicts the clinical severity of PSP at the individual level (adjusted R2: 0.739 and 0.892, p < 0.001). The machine learning algorithm for selection of diffusion tensor imaging-based features is accurate in predicting motor subscale of unified Parkinson’s disease rating scale and postural instability and gait disturbance of PSP.

Download Full-text

Prediction of Aphasia Severity in Patients with Stroke Using Diffusion Tensor Imaging

Brain Sciences ◽

10.3390/brainsci11030304 ◽

2021 ◽

Vol 11 (3) ◽

pp. 304

Author(s):

Jin-Kook Lee ◽

Myoung-Hwan Ko ◽

Sung-Hee Park ◽

Gi-Wook Kim

Keyword(s):

Diffusion Tensor Imaging ◽

White Matter ◽

Fractional Anisotropy ◽

Language Impairment ◽

Operating Characteristic ◽

Diffusion Tensor ◽

Characteristic Curve ◽

Verbal Comprehension ◽

Operating Characteristic Curve ◽

The Individual

This study classified the severity of aphasia through the Western Aphasia Battery and determined the optimal cut-off value for each Language-Related White Matter fiber and their combinations, we further examined the correlations between Language-Related White Matter and Western Aphasia Battery subscores. This retrospective study recruited 64 patients with aphasia. Mild/moderate and severe aphasia were classified according to cut-off Aphasia Quotient score of 51 points. Diffusion tensor imaging and fractional anisotropy reconstructed Language-Related White Matter in multiple fasciculi. We determined the area under the covariate-adjusted receiver operating characteristic curve to evaluate the accuracy of predicting aphasia severity. The optimal fractional-anisotropy cut-off values for the individual fibers of the Language-Related White Matter and their combinations were determined. Their correlations with Western Aphasia Battery subscores were analyzed. The arcuate and superior longitudinal fasciculi showed fair accuracy, the inferior frontal occipital fasciculus poor accuracy, and their combinations fair accuracy. Correlations between Language-Related White Matter parameters and Western Aphasia Battery subscores were found between the arcuate, superior longitudinal, and inferior frontal occipital fasciculi and spontaneous speech, auditory verbal comprehension, repetition, and naming. Diffusion-tensor-imaging-based language-Related White Matter analysis may help predict the severity of language impairment in patients with aphasia following stroke.

Download Full-text

Diffusion Tensor Imaging Reveals White Matter Changes Associated With Cognitive Status in Patients With Parkinson’s Disease

American Journal of Alzheimer s Disease & Other Dementias® ◽

10.1177/1533317512470207 ◽

2012 ◽

Vol 28 (2) ◽

pp. 154-164 ◽

Cited By ~ 33

Author(s):

Bingmei Deng ◽

Yuhu Zhang ◽

Lijuan Wang ◽

Kairun Peng ◽

Lixin Han ◽

...

Keyword(s):

Parkinson’S Disease ◽

Parkinson's Disease ◽

Cognitive Impairment ◽

Diffusion Tensor Imaging ◽

White Matter ◽

Rating Scale ◽

Diffusion Tensor ◽

Region Of Interest ◽

Cognitive Status ◽

Cerebral White Matter

Objective : Cognitive deficit and white matter alteration relationships in Parkinson’s disease (PD) were investigated. Methods : Comparison of 64 patients with PD (M:F, 34:30; 64.4 ± 10.4 years) classified as cognitively normal (PD-CogNL, n = 24), mild cognitive impairment (PD-MCI, n = 30), and dementia (PD-D, n = 10) with 21 healthy participants (M:F, 10:11; 60.1 ± 13.6 years) was conducted using white matter fractional anisotropy (FA), region-of-interest analysis, and diffusion tensor imaging. Results : The PD-D and PD-MCI exhibited higher Unified Parkinson’s Disease Rating Scale motor scores ( P < .001; P < .01) and Hoehn-Yahr stages ( P < .001; P < .05) and FA reductions in left frontal/right temporal white matter and bilateral anterior cingulated bundles. Largest FA reductions occurred in PD-D left anterior cingulated bundle and corpus callosum splenium. Disease durations of PD-D = 6.8 ± 6.86, PD-MCI = 5.1 ± 2.9, and PD-CogNL = 4.7 ± 3.4 years, suggesting progressive deterioration. Conclusions : Cerebral white matter deterioration may underlie progressive cognitive impairment in PD.

Download Full-text