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

Accumulating diffusion tensor imaging evidence suggests that white matter abnormalities evaluated by local diffusion homogeneity or fractional anisotropy occur in patients with blepharospasm, both of which are significantly correlated with disease severity. However, whether the individual severity of blepharospasm can be identified using these diffusion tensor imaging metrics remains unknown. We aimed to investigate whether a combination of machine learning techniques and local diffusion homogeneity or fractional anisotropy can accurately identify the individual severity of blepharospasm. Forty-one patients with blepharospasm were assessed using the Jankovic Rating Scale and diffusion tensor imaging. 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 local diffusion homogeneity or fractional anisotropy in 68 white matter regions. The proposed machine learning scheme with local diffusion homogeneity or fractional anisotropy 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 local diffusion homogeneity or fractional anisotropy measurements and a sophisticated machine learning scheme can accurately and reliably identify the individual disease severity in patients with blepharospasm.