Researchers at Florida Atlantic University (FAU) have developed a deep learning model that uses artificial intelligence to analyze electroencephalography (EEG) signals, enabling accurate detection and evaluation of dementia type and severity. The method offers a faster, noninvasive alternative to traditional imaging techniques such as MRI and PET scans, which are often expensive, time-consuming, and require specialized equipment.
Dementia affects millions in the United States, with Alzheimer’s disease (AD) being the most common form. In 2025, approximately 7.2 million Americans aged 65 and older are expected to have AD. Frontotemporal dementia (FTD), though less common overall, is the second leading cause of early-onset dementia, frequently impacting individuals between their 40s and 60s. While both conditions damage the brain differently—AD primarily impairs memory and spatial awareness while FTD targets behavior, personality, and language—their overlapping symptoms can lead to misdiagnosis.
Traditional diagnostic methods using EEG present challenges due to noisy signals and variability between individuals. Previous machine learning approaches yielded inconsistent results in differentiating AD from FTD.
The FAU research team addressed these issues by creating a deep learning model capable of analyzing both frequency- and time-based patterns in brain activity associated with each disease. Their study was published in Biomedical Signal Processing and Control.
Findings showed that slow delta brain waves were key biomarkers for both AD and FTD, especially in the frontal and central regions of the brain. AD was associated with more widespread disruption across additional regions and frequency bands like beta waves, suggesting greater overall brain damage compared to FTD.
The model achieved over 90% accuracy in distinguishing people with dementia from cognitively normal participants. It also predicted disease severity with relative errors below 35% for AD and 15.5% for FTD.
To improve specificity—the ability to correctly identify those without disease—the researchers used feature selection methods that increased this measure from 26% to 65%. Their two-stage approach first identified healthy individuals before separating cases of AD from FTD; this process resulted in an accuracy rate of 84%, ranking it among the top EEG-based diagnostic tools currently available.
Tuan Vo, first author of the study and doctoral student at FAU’s Department of Electrical Engineering and Computer Science stated: “What makes our study novel is how we used deep learning to extract both spatial and temporal information from EEG signals. By doing this, we can detect subtle brainwave patterns linked to Alzheimer’s and frontotemporal dementia that would otherwise go unnoticed. Our model doesn’t just identify the disease – it also estimates how severe it is, offering a more complete picture of each patient’s condition.”
Additional insights indicated that Alzheimer’s typically has broader effects on cognitive function than FTD—a finding consistent with previous neuroimaging studies but newly observed through inexpensive EEG data analysis.
Hanqi Zhuang, Ph.D., co-author as well as associate dean and professor at FAU’s Department of Electrical Engineering and Computer Science added: “Our findings show that Alzheimer’s disease disrupts brain activity more broadly, especially in the frontal, parietal and temporal regions, while frontotemporal dementia mainly affects the frontal and central areas. This difference explains why Alzheimer’s is often easier to detect. However, our work also shows that careful feature selection can significantly improve how well we distinguish FTD from Alzheimer’s.”
The new system integrates convolutional neural networks with attention-based LSTMs for detecting both type and severity of dementia using EEG data alone; visualization tools help clinicians interpret which signals influenced AI decisions.
Stella Batalama, Ph.D., dean of FAU’s College of Engineering and Computer Science said: “This work demonstrates how merging engineering, AI and neuroscience can transform how we confront major health challenges. With millions affected by Alzheimer’s and frontotemporal dementia, breakthroughs like this open the door to earlier detection, more personalized care, and interventions that can truly improve lives.”
Other contributors included Ali K. Ibrahim, Ph.D., assistant professor of teaching; Chiron Bang; all affiliated with FAU’s Department of Electrical Engineering & Computer Science.
