Summary: New research shows that deep learning can use EEG signals to distinguish Alzheimer's disease from Frontotemporal dementia with high accuracy. By analyzing both the time and the frequency of brain activity, the model uncovered different patterns: extensive disturbances in other areas in many areas of Alzheimer's and early local changes made in the area of dementia.
The system also measured the severity of the disease, giving doctors faster insight than traditional tools. These findings suggest that low-cost technology, paired with advanced AI, could enable personalized diagnostics and care for people to lose weight.
Basic facts
- Eg biomasker: Slow Delta waves in the middle and middle regions signal diseases in both cases.
- Different patterns: Alzheimer's has shown to be a widespread disorder, while prototemporal dementia remains localized.
- High accuracy: The deep learning system achieved 84% accuracy in classifying the two problems.
Source: Dude
Dementia is a group of disorders that gradually impair memory, thinking and daily functioning. Alzheimer's disease (ad), the most common form of dementia, will affect approximately 7.2 million Americans aged 65 and older by 2025.
Frontotemporal Dementia (FTD), while RARER, is the second most common cause of early onset dementia, which usually strikes people in their 40s to 60s.
Although both diseases damage the brain, they do so in different ways. Advertising mainly affects memory and spatial awareness, while FTD targets regions responsible for character, personality and language.
Because their symptoms can pass, it often leads to misdiagnosis. Differentiating between them is not only a scientific challenge but a clinical necessity, as an accurate diagnosis can greatly affect treatment, care and quality of life.
MRI and PET SCANS are useful for ad screening but are expensive, time consuming and require special equipment. Electroencephalography (EEG) offers a portable, non-invasive and inexpensive method for measuring brain activity with sensors across various normal bands.
However, the signals are often noisy and vary between individuals, making analysis difficult. Even with Machine Learning application to EEG EEG data, the results are conflicting and distinguishing AD from FTD remains difficult.
To address this issue, researchers from the College of Engineering and Computer Science at Florida Atlantic University developed a deep learning model that detects and evaluates advertising and FTD. It increases the accuracy of EEG and studies by analyzing both frequency- and time-based brain activity linked to each disease.
The results of the study, published in the journal Signal processing and biomedical control, It was found that slow Delta brain waves were an important biomarker for both AD and FTD, especially in the middle and upper regions of the brain.
With ads, brain activity was widely disrupted, and it affects other brain regions and normal bands such as beta, indicating extensive brain damage. This difference helps explain why the ad is easier to see in FTD.
The model achieved more than that 90% accuracy in distinguishing individuals with dementia (AD or FTD) from normal cognitive participants. It also predicted disease severity with a relative error of less than 35% for 15.5% of FTD.
Because AD and FTD share similar symptoms and brain function, self-report is difficult. Using Feature Selection, the researchers increased the model's specificity – how well it identified people without the disease – from 26% to 65%.
Their two-stage design – first to find healthy people, then to distinguish AD from FTD – achieved an accuracy of 84%, a position among the best EEG methods to date.
The model combines convolutional neural networks and attention-based LSSMs to detect the type and severity of dementia from EEG data. GRED-CAM shows which brain signals influence the model, helping medical doctors understand its decisions.
This approach provides a new perspective on how brain activity occurs and which regions and frequencies drive it – something traditional tools rarely capture.
“What makes our study novel is how we used deep learning to extract both spatial and temporal information from EEG signals,” said Tuan VO, first author of FAU's Department of Engineering.
“By doing this, we can find subtle brainwave patterns linked to Alzheimer's and Alzheimer's dementia and Frontotemporal dementia that have been overlooked in the disease – and measure how severe it is, giving a complete picture of each patient's condition.”
The findings also revealed that the ad tends to be more severe, has a wider impact on brain areas and leads to lower mental schools, while the effects of FTD are more organized in the frontal and temporal lobes.
This understanding is consistent with previous studies of neuroimaging research but adds new depth by showing how these markers appeared in the EG Tool – A cheap and inexpensive diagnostic tool.
“Our findings show that Alzheimer's disease affects brain function widely, especially in the frontal areas, Pang said.
“The difference explains why Alzheimer's is so easy to detect. However, our work also shows that careful trait selection can greatly improve how we diagnose FTD from Alzheimer's.”
Overall, research shows that deep learning can advance the diagnosis of dementia by integrating detection and assessment of severity into a single system, cutting out lengthy tools and providing real-time tools to track disease progression.
“This work shows how combining engineering, AI and neuroscience can revolutionize how we approach major health challenges,” said Stella Batalama, Ph.D., Dean of the College of Engineering and Computer Science.
“With billions affected by Alzheimer's and frontotemporal dementia, breakthroughs like this open the door to earlier detection, more personalized care, and interventions that can improve real lives.”
Study authors are Ali K. Ibrahim, Ph.D., professor of education; And Chiron Bang, doctoral student, both with FAU's electrical engineering and computer science department.
Important Questions Answered:
A: Their symptoms and eeg signatures tend to overlap, leading to misdiagnosis without special considerations.
A: It analyzes spatial and temporal characteristics simultaneously, revealing subtle differences in brainwaves that are lost in conventional methods.
A: Yes – It measures the height levels in both cases, helping medical doctors to continue the organization more effectively.
Editing notes:
- This article was edited by the editor of neuroscience news.
- The journal is fully reviewed.
- Additional context added by our staff.
About This Ai and Neurotech News News
Author: Gisele Galous
Source: Dude
Contact: Gisele Galous – Fau
Image: This photo is posted in Neuroscience News
Actual research: Open access.
“Extraction and interpretation of EEG features for diagnosis and over-prediction of Alzheimer's disease and total dementia using deep learning” by Tuan Vo et al. Biomedical Signal Processing and Control
-Catshangwa
Extraction and interpretation of EEG features for diagnosis and over-prediction of Alzheimer's disease and dementia using deep learning
Alzheimer's disease (ad) is the most common form of dementia, characterized by progressive cognitive decline and memory loss. Frontotemporal Dementia (FTD), the second most common form of dementia, affects the frontal and temporal lobes, causing changes in personality, behavior and cognition.
Because of the overwhelming number of symptoms, FTD is often misdiagnosed as an ad. Although electroencephalography (EEG) is portable, non-invasive and inexpensive, and less expensive, its diagnostic power for AD and FTD is limited by the similarities between the two diseases.
To address this, we present an EEG-based approach to detect and predict the magnitude of AD and FTD using deep learning. Key findings include the functions of the Delta Band in the anterior and posterior regions as a biomasker.
By extracting temporal and spectral features from the EEG signal, our model integrates a neural network based on short-term memory (Alstm), achieving more than 90% accuracy in distinguishing people with FTD from normal people (CN).
It also reports robustly with a range of errors of less than 35% for AD and around 15.5% for FTD. Differentiating FTD from AD remains challenging due to shared symptoms.
However, using the feature selection process improves the specificity of distinguishing AD from FTD, increasing it from 26% to 65%. Building on this, we developed a two-stage method to classify AD, CN, and FTD simultaneously. In this approach, CN is identified first, followed by the differentiation of FTD from AD.
This method achieves an overall accuracy of 84% in classifying AD, CN, and FTD.
