EEG-based Neglect Detection for Stroke Patients.

Area of Science:

• Neuroscience
• Neurology
• Biomedical Engineering

Background:

• Spatial neglect (SN) is a common neurological deficit after stroke, characterized by inattention to the contralesional visual field.
• Current assessment methods like the Behavioral Inattention Test (BIT) have limitations, including unreliability, limited scope, and lack of ecological validity.
• There is a need for more objective, sensitive, and dynamic methods to assess SN and its severity.

Purpose of the Study:

• To develop and validate an electroencephalography-based brain-computer interface (EEG-BCI) for detecting spatial neglect (SN) in stroke patients.
• To utilize the Starry Night Test within an EEG-BCI framework to overcome the limitations of traditional SN assessment tools.
• To accurately distinguish between stroke patients with and without SN, and to lay the groundwork for assessing neglect severity.

Main Methods:

• Implementation of an electroencephalography (EEG)-based brain-computer interface (BCI).
• Utilized the Starry Night Test as a dynamic assessment tool within the BCI.
• Employed a convolutional neural network (CNN) model for analyzing EEG data to detect SN.

Main Results:

• The proposed EEG-BCI system demonstrated high accuracy, specificity, and sensitivity in detecting spatial neglect.
• The system successfully distinguished between stroke patients with and without neglect.
• The CNN-based analysis of EEG data proved effective for neglect detection.

Conclusions:

• The EEG-BCI utilizing the Starry Night Test offers a promising, accurate, and sensitive method for spatial neglect detection in stroke survivors.
• This technology addresses the limitations of conventional pen-and-paper assessments.
• Future research aims to refine the system for estimating the field of view (FOV) for a more comprehensive SN assessment.