Can Children With Perinatal Stroke Use a Simple Brain Computer Interface?. Issue 7 (27th May 2021)

• Record Type: Journal Article
• Authors: Jadavji, Zeanna; Zhang, Jack; Paffrath, Brett; Zewdie, Ephrem; Kirton, Adam
• Is Part Of: Stroke. Volume 52:Issue 7(2021)
• Page Start: 2363
• Page End: 2370
• Publication Date: 2021-05-27
• Subjects: brain -- brain computer interface -- cerebral palsy -- electroencephalography magnetic resonance imaging; Cerebrovascular disease -- Periodicals; Cerebral circulation -- Periodicals; 616.81
• DOI: 10.1161/STROKEAHA.120.030596
• Languages: English
• ISSNs: 0039-2499

Abstract

Abstract : Supplemental Digital Content is available in the text. Abstract : Background and Purpose: Perinatal stroke is the leading cause of hemiparetic cerebral palsy resulting in lifelong disability for millions of people worldwide. Options for motor rehabilitation are limited, especially for the most severely affected children. Brain computer interfaces (BCIs) sample brain activity to allow users to control external devices. Functional electrical stimulation enhances motor recovery after stroke, and BCI-activated functional electrical stimulation was recently shown to improve upper extremity function in adult stroke. We aimed to determine the ability of children with perinatal stroke to operate a simple BCI. Methods: Twenty-one children with magnetic resonance imaging–confirmed perinatal stroke (57% male, mean [SD] 13.5 [2.6] years, range 9–18) were compared with 24 typically developing controls (71% male, mean age [SD] 13.7 [3.7] years, range 6–18). Participants trained on a simple EEG-based BCI over 2 sessions (10 trials each) utilizing 2 different mental imagery strategies: (1) motor imagery (imagine opening and closing of hands) and (2) goal oriented (imagine effector object moving toward target) to complete 2 tasks: (1) drive a remote controlled car to a target and (2) move a computer cursor to a target. Primary outcome was Cohen Kappa with a score >0.40 suggesting BCI competence. Results: BCI performance was comparable between stroke and control participants. Mean scores were 0.39 (0.18) for stroke versus 0.42 (0.18) for controls (t[42]=0.478, P =0.94). No difference in performance between venous (M=0.45, SD=0.29) and arterial (M=0.34, SD=0.22) stroke (t[82]=1.89, P =0.090) was observed. No effect of task or strategy was observed in the stroke participants. Over 90% of stroke participants demonstrated competency on at least one of the 4 task-strategy combinations. Conclusions: Children with perinatal stroke can achieve proficiency in basic tasks using simple BCI systems. Future directions include exploration of BCI-functional electrical stimulation systems for rehabilitation for children with hemiparesis and other forms of cerebral palsy.