Optimizing Bioimaging: Quantum Computing-Inspired Bald Eagle Search Optimization for Motor Imaging EEG Feature Selection

One of the most important objectives in brain-computer interfaces (BCI) is to identify a subset of characteristics that represents the electroencephalographic (EEG) signal while eliminating elements that are duplicate or irrelevant. Neuroscientific research is advanced by bioimaging, especially in the field of BCI. In this work, a novel quantum computing-inspired bald eagle search optimization (QC-IBESO) method is used to improve the effectiveness of motor imagery EEG feature selection. This method can prevent the dimensionality curse and improve the classification accuracy of the system by lowering the dimensionality of the dataset. The dataset that was used in the assessment is from BCI Competition-III IV-A. To normalize the EEG data, Z-score normalization is used in the preprocessing stage. Principal component analysis reduces dimensionality and preserves important information during feature extraction. In the context of motor imagery, the QC-IBESO approach is utilized to select certain EEG characteristics for bioimaging. This facilitates the exploration of intricate search spaces and improves the detection of critical EEG signals related to motor imagery. The study contrasts the suggested approach with conventional methods like neural networks, support vector machines and logistic regression. To evaluate the efficacy of the suggested strategy in contrast to current techniques, performance measures such as F1-score, precision, accuracy and recall are computed. This work advances the field of feature selection techniques in bioimaging and opens up a novel and intriguing direction for the investigation of quantum-inspired optimization in neuroimaging.