Developing an EEG-fNIRS Neurofeedback Application for Brain Training for Autistic Children

Autism spectrum disorder (ASD) is a lifelong neurodevelopmental condition characterized by difficulties in social communication and interaction, often accompanied by cognitive and emotional regulation challenges. In Hong Kong and many other countries, ASD is increasingly prevalent. Despite this, the brain health of autistic individuals has been relatively neglected in both healthcare systems and public policies. There is also a lack of approaches and technologies that directly intervene with brain function. Since many autistic children experience poor vocational and health outcomes in adulthood, there is a strong need to develop effective and accessible neuroscience-based treatments.

This project aims to apply cutting-edge neuroscientific methods to develop an innovative closed-loop brain training intervention for children with ASD. The intervention will combine electroencephalography (EEG) and functional near-infrared spectroscopy (fNIRS) in a unified neurofeedback training system. Neurofeedback training teaches individuals to self-regulate brain activity by providing real-time feedback. In the traditional neurofeedback study, EEG has been used to guide neurofeedback by monitoring electrical activity in the brain, while more recently fNIRS has been used to track hemodynamic activity. However, no existing neurofeedback system has integrated these two modalities. Combining EEG and fNIRS provides an opportunity to enhance neurovascular coupling, the relationship between neural activity and blood flow, which is often altered in neuropsychiatric conditions such as autism.

The proposed neurofeedback application will include multiple training modules designed to address cognitive, emotional, and social difficulties common in autism. The cognitive training module will target brain activity patterns associated with attention and executive function. The affective training module will focus on modulating frontal brain activity linked to emotional regulation. The social training module will aim to enhance neural and hemodynamic activity associated with social cognition and communication. By integrating both EEG and fNIRS indices, the system will encourage children to regulate electrical and hemodynamic activity simultaneously, which cannot be achieved using either modality alone.

To maximize engagement, the application will incorporate ecologically valid feedback stimuli and reward-based learning principles. Instead of relying solely on abstract indicators such as bars or tones, the feedback will involve intrinsically rewarding stimuli, such as videos or positive visual cues, to increase motivation and adherence. The training difficulty will be adjusted progressively based on individual performance to ensure sustained engagement and improvement.

In addition, the system will be developed as a cross-device application using open-source lab streaming layer (LSL) software, ensuring compatibility with a wide range of EEG and fNIRS devices. The hardware and software will be optimized to ensure high-quality signals, including the use of shielded wet electrodes for EEG to reduce noise and short-separation channels in fNIRS to minimize extracerebral signal contamination. These features will allow neurofeedback training to be conducted with minimal environmental interference, enhancing both reliability and clinical applicability.