Basic Science and Pathogenesis

Area of Science:

• Neuroscience
• Biomedical Engineering
• Cognitive Science

Background:

• Dementia affects over 55 million globally, with limited success from traditional interventions.
• Alzheimer's disease constitutes 60-70% of dementia cases.
• Brain-Machine Interfaces (BMIs) offer a novel approach for early detection and rehabilitation.

Purpose of the Study:

• Investigate the feasibility of non-invasive BMIs for monitoring cognitive health in dementia patients.
• Explore BMIs for early dementia detection using neural biomarkers.
• Assess BMIs for personalized dementia rehabilitation and patient engagement.

Main Methods:

• Utilize non-invasive sensors (EEG, fNIRS) to monitor dementia-related neural biomarkers like theta-gamma oscillations and P300.
• Employ machine learning algorithms for early dementia risk classification.
• Implement real-time neurofeedback and gamified interfaces for cognitive rehabilitation.

Main Results:

• EEG and machine learning classifiers show promising results for dementia detection.
• A BiLSTM model achieved 97.27% accuracy in detecting dementia from EEG signals.
• Passive monitoring via wearable BMIs can detect anomalies during daily activities.

Conclusions:

• BMIs hold transformative potential for dementia care, integrating early detection with personalized rehabilitation.
• BMIs can identify at-risk individuals years before symptom onset.
• Adaptive neurofeedback systems may slow cognitive decline by strengthening cognitive reserve.