Swedish Researchers Develop Advanced AI Models for Early Dementia Diagnosis Using EEG

Researchers at Örebro University have made significant strides in dementia diagnosis through the development of two advanced AI models that analyze brain electrical activity via EEG data. These innovations aim to improve early detection of Alzheimer's disease and frontotemporal dementia, offering the potential to slow disease progression and enhance patients' quality of life.

The first AI model, detailed in a study titled "An explainable and efficient deep learning framework for EEG-based diagnosis of Alzheimer’s disease and frontotemporal dementia," employs temporal convolutional networks and LSTM networks. It analyzes EEG signals by breaking them down into frequency bands and achieves over 80% accuracy in distinguishing between dementia types and healthy controls. Importantly, this AI integrates explainable AI techniques, allowing clinicians to understand how diagnoses are reached, thus reducing the 'black box' concern.

Complementing this, the second study, "Privacy–preserving dementia classification from EEG via hybrid–fusion EEGNetv4 and federated learning," presents a privacy-conscious AI model under one megabyte designed to protect patient data. Through federated learning, this model enables multiple healthcare centers to collaboratively train the AI without exchanging sensitive information, achieving more than 97% accuracy.

Muhammad Hanif, an informatics researcher at Örebro University, emphasized that early diagnosis is crucial for proactive interventions. The accessibility of EEG as a simple, cost-effective measure already used in healthcare suggests these AI tools could soon broaden diagnostic reach, potentially even enabling at-home testing in the future.

Looking ahead, the research team plans to expand their studies with larger and more diverse datasets, explore additional EEG characteristics, and extend research into other dementia forms, all while upholding strict data privacy and security.

This breakthrough aligns with broader observations from Lund University, where older adults demonstrate openness to user-friendly and meaningful AI technologies, particularly those enhancing health and independence. Such receptiveness is vital for integrating advanced AI diagnostic tools into elder care effectively.