Artificial Deep Learning Brain-Actuated Lower Limb Exoskeleton for Paralyzed

Stroke-induced paralysis often leaves patients dependent on caregivers for daily activities. While several assistive technologies have been developed for post-stroke rehabilitation, most lack effective user control, sensory feedback, or fall-detection capabilities, limiting their practical utility. Conventional rehabilitation remains essential but insufficient for full functional recovery. The Brain–Computer Interface (BCI) presents a promising approach for restoring motor function; however, existing systems often face challenges such as signal variability, high error rates, and user fatigue caused by cumbersome exoskeletons. In Phase 1, we address these limitations through a Brain-Controlled Lower Limb Exoskeleton (BCLLE) that enables movement based on user intentions decoded from EEG signals. The system incorporates an adaptive feedback mechanism to minimize false activations and a flexible, anatomically accurate carbon-fiber design adaptable to different degrees of paralysis. Using the Novel-T Symmetric Encryption Algorithm (NTSA), the exoskeleton securely transmits user status and emergency alerts to caregivers. The BCLLE achieved a classification accuracy exceeding 80%, outperforming comparable systems. In Phase 2, we introduce the Artificial Muscle Intelligence with Deep Learning (AMIDL) system, which eliminates the need for a physical exoskeleton by integrating EEG-based intent recognition with Transcutaneous Electrical Nerve Stimulation (TENS). This enables upper limb movements directly through artificial muscle activation. The system also employs gesture recognition to convert detected gestures into voice commands, facilitating communication for paralyzed individuals. Experimental validation on healthy and paralyzed subjects demonstrated that AMIDL reduces mental fatigue, frustration, and operational errors while maintaining continuous, real-time control. The use of lightweight wireless electrodes further minimizes physical strain. Overall, the proposed BCLLE–AMIDL framework enhances motor rehabilitation and communication, improving autonomy and quality of life for individuals with paralysis.