Engineers Develop AI Model to Differentiate Dementia Types
Researchers at Florida Atlantic University have made a significant advancement in the diagnosis of dementia types using artificial intelligence and EEG brainwave analysis. Their newly developed deep learning model can effectively distinguish between Alzheimer’s Disease (AD) and Frontotemporal Dementia (FTD), which are both leading causes of cognitive impairment but present differently in terms of symptoms and brain activity.
Dementia encompasses various disorders that progressively hinder memory, cognitive function, and daily activities. By 2025, approximately 7.2 million Americans aged 65 and older are projected to be living with Alzheimer’s, the most prevalent form of dementia. In contrast, FTD, while less common, serves as the second leading cause of early-onset dementia, often affecting individuals between their 40s and 60s. The overlapping symptoms of these two conditions often lead to misdiagnosis, which can adversely affect treatment outcomes and quality of life.
Innovative Approach to Diagnosis
Traditional diagnostic methods such as MRI and PET scans, while effective, can be costly, time-consuming, and require specialized facilities. In comparison, Electroencephalography (EEG) is a more accessible, noninvasive alternative that measures brain activity through external sensors. However, the analysis of EEG data has been complicated by noisy signals and individual variability, making accurate differentiation between AD and FTD a challenge.
The research team from the College of Engineering and Computer Science at Florida Atlantic University addressed these challenges by developing a deep learning model that enhances the accuracy of EEG readings. Their study, published in the journal Biomedical Signal Processing and Control, revealed that slow delta brain waves serve as a key biomarker for both Alzheimer’s and FTD, particularly in the frontal and central brain regions.
The findings indicate that Alzheimer’s disrupts brain activity more extensively than FTD, affecting additional brain regions and frequency bands such as beta. This broader disruption contributes to the relative ease of diagnosing Alzheimer’s compared to FTD, which mainly impacts frontal and temporal lobes.
Promising Results and Future Implications
The new model achieved over 90% accuracy in distinguishing individuals with dementia from cognitively normal participants. It also predicted disease severity with relative errors of less than 35% for Alzheimer’s and 15.5% for FTD. By employing feature selection techniques, the researchers improved the model’s specificity from 26% to 65%, enhancing its ability to correctly identify healthy individuals.
The two-stage design of the model first detects healthy participants and then differentiates between AD and FTD, achieving an accuracy rate of 84%. This performance ranks among the best EEG-based methods developed to date. The model utilizes convolutional neural networks combined with attention-based long short-term memory (LSTM) networks to assess both the type and severity of dementia from EEG data. Additionally, the Grad-CAM visualization technique highlights which brain signals influenced the model’s decisions, providing valuable insights for clinicians.
Tuan Vo, the study’s first author and a doctoral student at FAU, emphasized the novelty of their approach. He 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.”
The results also align with existing neuroimaging studies, revealing that Alzheimer’s typically manifests as a more severe condition, affecting a broader range of brain areas and resulting in lower cognitive scores. In contrast, the effects of FTD are more localized.
Dr. Hanqi Zhuang, co-author and associate dean at FAU, remarked, “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.”
The implications of this research extend beyond diagnosis. By merging engineering, artificial intelligence, and neuroscience, the study presents a streamlined approach to dementia diagnosis that can significantly reduce evaluation times and provide clinicians with real-time tools to monitor disease progression.
Stella Batalama, dean of the College of Engineering and Computer Science, concluded, “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.”
The research team also included Ali K. Ibrahim, Ph.D., an assistant professor of teaching, and Chiron Bang, a doctoral student, both from the FAU Department of Electrical Engineering and Computer Science. For further details, refer to the study titled “Extraction and interpretation of EEG features for diagnosis and severity prediction of Alzheimer’s Disease and Frontotemporal dementia using deep learning,” published in Biomedical Signal Processing and Control in 2026.
