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

Abstract

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.

Cortico-Subcortical White Matter Bundle Changes in Cervical Dystonia and Blepharospasm

“…The results of the present study showing changes in specific white matter tracts expand previous literature on CD and BSP, demonstrating diffuse microstructural damage in the basal ganglia, cerebellum and sensorimotor cortical areas [ 10 , 11 , 12 , 13 ], as well as in the white matter [ 11 , 19 , 20 , 51 , 52 ]. Moreover, the WM tracts we reconstructed correspond to brain regions with microstructural integrity loss in the WM adjacent to the pallidum and putamen and the precentral and postcentral gyri [ 13 , 17 , 53 ].…”

“…Microstructural alterations were also found in the white matter (WM) adjacent to the primary sensorimotor, inferior parietal and middle cingulate cortices in patients with CD [ 13 , 14 , 15 ] and BSP [ 16 , 17 , 18 ]. Finally, studies with whole-brain approaches showed WM microstructural disruption in the corpus callosum, the internal capsule and the white matter underlying the sensorimotor cortex in CD and BSP patients [ 11 , 19 , 20 ].…”

Detection of mild cognitive impairment in Parkinson’s disease using gradient boosting decision tree models based on multilevel DTI indices

“…LDH is independent of the diffusion model and can reveal intervoxel diffusion properties by capturing the overall coherence of water molecule diffusion within a neighbourhood; it reflects the microstructural coherence of the underlying WM fibres and provides additional insights beyond traditional intravoxel metrics. Some recent studies have applied LDH parameters to the detection of WM microstructure abnormalities in vascular cognitive impairment [ 15 ], epilepsy [ 16 ], type 2 diabetes [ 17 ], and blepharospasm [ 18 ], demonstrating regions of variation that differed from those of intravoxel diffusion metrics. Moreover, LDH can help predict the prognosis of stroke patients [ 19 ].…”