Machine Learning-Based Prediction of Changes in Behavioral Outcomes Using Functional Connectivity and Clinical Measures in Brain-Computer Interface Stroke Rehabilitation
The goal of this work is to evaluate if changes in brain connectivity can predict behavioral changes among subjects who have suffered stroke and have completed brain-computer interface (BCI) interventional therapy. A total of 23 stroke subjects, with persistent upper-extremity motor deficits, received the stroke rehabilitation therapy using a closed-loop neurofeedback BCI device. Over the course of the entire interventional therapy, resting-state fMRI were collected at two time points: prior to start and immediately upon completion of therapy. Behavioral assessments were administered at each time point via neuropsychological testing to collect measures on Action Research Arm Test, Nine-Hole Peg Test, Barthel Index and Stroke Impact Scale. Resting-state functional connectivity changes in the motor network were computed from pre- to post-interventional therapy and were combined with clinical data corresponding to each subject to estimate the change in behavioral performance between the two time-points using a machine learning based predictive model. Inter-hemispheric correlations emerged as stronger predictors of changes across multiple behavioral measures in comparison to intra-hemispheric links. Additionally, age predicted behavioral changes better than other clinical variables such as gender, pre-stroke handedness, etc. Machine learning model serves as a valuable tool in predicting BCI therapy-induced behavioral changes on the basis of functional connectivity and clinical data.
KeywordsBrain-computer interface Stroke rehabilitation BCI therapy Upper extremity motor recovery Resting-state fMRI Machine learning Predictive model
The authors would like to thank all the subjects and their families for their participation in the stroke rehabilitation program. Thanks to the study coordinator Theresa Jungae Kang and the MRI technicians Sara John and Jenelle Fuller at the Wisconsin Institutes for Medical Research. This study was supported by NIH grants RC1MH090912-01, K23NS086852, T32GM008692, UL1TR000427, T32EB011434 and TL1TR000429. Additional funding was also provided through a Coulter Translational Research Award, an American Heart Association Postdoctoral Fellow Research Award, AHA Midwest Grant-in-Aid Award, AHA National Innovation Award, UW Milwaukee-Madison Intercampus Grants, UW Graduate School, Grants from Shapiro Foundation and Foundation of ASNR award.
- 1.Bajaj, S., Butler, A.J., Drake, D., Dhamala, M.: Functional organization and restoration of the brain motor-execution network after stroke and rehabilitation. Front. Hum. Neurosci. 9(173), 1–14 (2015)Google Scholar
- 2.Carter, A.R., Astafiev, S.V., Lang, C.E., Connor, L.T., Rengachary, J., Strube, M.J., Pope, D.L.W., Shulman, G.L., Corbetta, M.: Resting inter-hemispheric fMRI connectivity predicts performance after stroke. Ann. Neurol. 67(3), 365–375 (2010)Google Scholar
- 8.Nair, V.A., Young, B.M., Nigogosyan, Z., Remsick, A., Weber, S., Diffee, K., Walton, L., Tyler, M., Sattin, J., Edwards, D.F., Williams, J., Prabhakaran, V.: Abstract 6: resting-state functional connectivity changes after stroke rehabilitation using closed loop neurofeedback. Stroke 4, A6 (2015)Google Scholar
- 11.Varkuti, B., Guan, C., Pan, Y., Phua, K.S., Ang, K.K., Kuah, C.W.K., Chua, K., Ang, B.T., Birbaumer, N., Sitaram, R.: Resting state changes in functional connectivity correlate with movement recovery for BCI and robot-assisted upper-extremity training after stroke. Neurorehabil. Neural Repair 27(1), 53–62 (2013)CrossRefGoogle Scholar
- 13.Young, B.M., Nigogosyan, Z., Walton, L.M., Song, J., Nair, V.A., Grogan, S.W., Tyler, M.E., Edwards, D.F., Caldera, K., Sattin, J.A., Williams, J.C., Prabhakaran, V.: Changes in functional brain organization and behavioral correlations after rehabilitative therapy using a brain-computer interface. Front. Neuroeng. 7(26), 1–15 (2014)Google Scholar
- 18.Mahoney, F.I., Barthel, D.: Functional evaluation: the Barthel Index. Maryland State Med J. 14, 56–61 (1965)Google Scholar
- 19.Scholkopf, B., Smola, A.J.: Learning with Kernels: Support Vector Machines, Regularization, Optimization, and Beyond. MIT Press, Cambridge, (2001)Google Scholar
- 23.Schölkopf, B., Smola, A.J.: Learning with Kernels: Support Vector Machines, Regularization, Optimization and Beyond, p. 633. MIT Press, Cambridge (2002)Google Scholar
- 24.Weston J., Elisseeff A., Bakir G., Sinz F.: The Spider Machine Learning Toolbox. Resource object oriented environment. http://people.kyb.tuebingen.mpg.de/spider/main.html