Identifying Neuroimaging Markers of Motor Disability in Acute Stroke by Machine Learning Techniques

Abstract

Conventional mass-univariate analyses have been previously used to test for group differences in neural signals. However, machine learning algorithms represent a multivariate decoding approach that may help to identify neuroimaging patterns associated with functional impairment in "individual" patients. We investigated whether fMRI allows classification of individual motor impairment after stroke using support vector machines (SVMs). Forty acute stroke patients and 20 control subjects underwent resting-state fMRI. Half of the patients showed significant impairment in hand motor function. Resting-state connectivity was computed by means of whole-brain correlations of seed time-courses in ipsilesional primary motor cortex (M1). Lesion location was identified using diffusion-weighted images. These features were used for linear SVM classification of unseen patients with respect to motor impairment. SVM results were compared with conventional mass-univariate analyses. Resting-state connectivity classified patients with hand motor deficits compared with controls and nonimpaired patients with 82.6-87.6% accuracy. Classification was driven by reduced interhemispheric M1 connectivity and enhanced connectivity between ipsilesional M1 and premotor areas. In contrast, lesion location provided only 50% sensitivity to classify impaired patients. Hence, resting-state fMRI reflects behavioral deficits more accurately than structural MRI. In conclusion, multivariate fMRI analyses offer the potential to serve as markers for endophenotypes of functional impairment.

Disruptions of network connectivity predict impairment in multiple behavioral domains after stroke

“…Second, alterations in interhemispheric connections showed the strongest association with behavioral impairment across nearly all domains. Reductions in interhemispheric coherence were predominant not only in the functional connectivity related to specific deficits (12,13,(28)(29)(30), but also in the multidomain FC that generalized across deficits (Fig. 6).…”

Functional resting-state connectivity of the human motor network: Differences between right- and left-handers

“…Afterwards, the classification model was optimized based on different soft margin constants of the linear separation boundary (i.e., C-parameters ranging from small (C = 0.0001) to large (C = 30). For this optimization, we used another leave-one-subject-out cross-validation procedure by successively omitting and classifying one subject of the training data (please see Rehme et al, 2014, for a more detailed description). We then computed the posterior balanced accuracy of classifications across all outer loops and reported 95% confidence intervals (CIs) (Brodersen et al, 2010).…”

“…To compare multivariate SVM classification with univariate classification, we also tested the classification accuracy of individual voxels within the motor network for the discrimination between right- and left-handers (Rehme et al, 2014). We performed the same leave-one-subject-out cross-validation procedure with a training sample and one independent test sample according to the multivariate SVM training procedure described above.…”

Combining Multiple Resting-State fMRI Features during Classification: Optimized Frameworks and Their Application to Nicotine Addiction

“…Thus, the resting-state approach is ideal to examine brain function in patients who may experience difficulty in performing tasks. Resting-state data are widely investigated using machine learning approaches (Deshpande et al, 2010; Shen et al, 2010; Dai et al, 2012; Eloyan et al, 2012; Zeng et al, 2014; Iidaka, 2015; Liu et al, 2015; Rehme et al, 2015). For example, our group previously applied SVM-based classification to resting-state functional connectivity (rsFC) data from 21 smokers and 21 non-smokers to successfully predict smoking status (Pariyadath et al, 2014).…”