Normalizing Go/NoGo ERP in the children with cerebral palsy using motor imagery-based BCI training

2048 

Abstract Submission 

Siao-Jhen Wu¹, Jeng-Ren Duann^2,3, Jen-Hua Chang⁴, Yueh-Ru Lee²

¹National Yang Ming Chiao Tung University, Hsinchu, Hsinchu, ²National Yang Ming Chiao Tung University, Hsinchu, Taiwan, ³Institute for Neural Computation, University of California San Diego, La Jolla, CA, ⁴Hemei Experimental School, Changhua, Taiwan

Siao-Jhen Wu  
National Yang Ming Chiao Tung University
Hsinchu, Hsinchu

Jeng-Ren Duann  
National Yang Ming Chiao Tung University|Institute for Neural Computation, University of California San Diego
Hsinchu, Taiwan|La Jolla, CA

Jen-Hua Chang  
Hemei Experimental School
Changhua, Taiwan

Yueh-Ru Lee  
National Yang Ming Chiao Tung University
Hsinchu, Taiwan

Cerebral palsy (CP) is a neurological disorder that affects movement, posture, and coordination, resulting from damage to the developing brain, often occurring before or during birth. This condition manifests in various forms, impacting muscle control and coordination, and is characterized by impaired motor function. In the context of the brain, cerebral palsy is associated with abnormalities in the motor cortex. The damage to this area disrupts the communication between the brain and muscles, leading to difficulties in coordinated movements and muscle control. The severity and specific manifestations of cerebral palsy vary among individuals, encompassing a spectrum of motor impairments that may affect different parts of the body. Early intervention and rehabilitative strategies play a crucial role in managing and improving the quality of life for individuals with CP. In this study, we employed a motor imagery-based brain-computer interface (BCI) EEG system, conjoined with a BCI-controlled robotic arm, to design BCI training games for children with CP. The objective was to improve the function of the primary motor cortex and alleviate muscle tones in the upper extremities of children with CP.

Sixteen children with CP, exhibiting varying degrees of severity, were screened by an occupational therapist at the Hemei Experimental School in Changhua, Taiwan, and participated in the study. The study employed an 8-week training protocol, conducted weekly with each session spanning one hour, and aimed at enhancing the modulation of the primary motor cortex in children with CP. In our study, a custom-made EEG spectrometer was used to capture the alpha suppression associated with the motor imagery processing as the feature to trigger a robotic arm in real-time to move a plastic ball along a designated track as the BCI-based training facility. After the 8-week training protocol, the effectiveness of the training was evaluated using a pre- and post-training Go/NoGo EEG experiment conducted using a 64-channel ANT EEG system (ANT Neuro, Hengelo, Netherlands) sampled at 1000 Hz. The EEG data were preprocessed and analyzed using EEGLAB (Delorme and Makeig, 2004) in the Matlab environment (MathWorks Inc., Boston, USA).

Results revealed that 10 of the 16 children exhibited the emergence of a N200 ERP component in the frontocentral midline region post-training, a phenomenon absent in the pre-training assessments. However, for those with severe CP, the BCI training failed to rectify the aberrant ERP patterns induced by the No/NoGo task.

This nuanced exploration sheds light on the potential efficacy of motor imagery-based BCI interventions in ameliorating motor-related neurophysiological signatures in children with CP and explores the feasibility of implementing this intervention. This work was supported in part by the National Science and Technoilogy Council, Taiwan (NSTC110-2511-H-A49 -012 -MY3 and NSTC110-2221-E-A49 -038).

Neurodevelopmental/ Early Life (eg. ADHD, autism)

Imagery ²

Brain Machine Interface ¹

Electroencephaolography (EEG)

Other - Cerebral Palsy, Brain-Computer Interface, ERP