This paper advances real-time cursor control for individuals with motor impairments through a novel brain–computer interface (BCI) system based solely on motor imagery. We introduce an enhanced deep neural network (DNN) classifier integrated with a Four-Class Iterative Filtering (FCIF) technique for efficient preprocessing of neural signals. The underlying approach is the Four-Class Filter Bank Common Spatial Pattern (FCFBCSP) and it utilizes a customized filter bank for robust feature extraction, thereby significantly improving signal quality and cursor control responsiveness. Extensive testing under varied conditions demonstrates that our system achieves an average classification accuracy of 89.1% and response times of 663 milliseconds, illustrating high precision in feature discrimination. Evaluations using metrics such as Recall, Precision, and F1-Score confirm the system’s effectiveness and accuracy in practical applications, making it a valuable tool for enhancing accessibility for individuals with motor disabilities.

Individuals exhibiting neurological, developmental and behavioral disorders can manifest stress, agitation and emotional deregulation. The prolonged emotional deregulation increases the development of depression, anxiety disorders and in extreme cases, this persistent deregulation contributes to onset of psychiatric symptoms. This study proposes a novel approach for the detection of psychiatric risk and abnormal behavioral states using multimodal biosensors integration namely Pulse, Electromyography (EMG), and Galvanic Skin Response (GSR) sensors. This experiment conducted in a skill-training centre for Endosulfan victims of age 13 ± 2 years with necessary care. Significant differences in the electro dermal signal such as GSR, pulse, EMG observed among the 66 subjects with multiple trails taken in this research work. Each subject data has multiple values of bio-signals. Data acquisition performed by integrating sensors onto the Arduino UNO microcontroller and Cool-Term software tool. In total, fifteen nonlinear transform domain features extracted. The confidence intervals of features verified using the t -test ( p < 0.05). Classification model created using the machine learning algorithms viz. Logistic Regression, Random Forests, Gradient Boosting, Support Vector Machines (SVM), K-Nearest Neighbors (KNN), XGBoost, and Neural Network. In which, Random Forest, XGBoost, and Neural Network achieved a sustainable accuracy of 97.92%. This approach of biosensor monitoring helps to predict consistently a behavioral state of differently abled individuals and supports health care providers in early identifications of behavioral and physiological markers indicative of psychiatric risks and sustainable mental health ecosystem.