New And Reliable Test to Predict Dementia Risk

Summary: The study developed a highly accurate EEG-based tool for early detection of MCI in older Black Americans, offering a soft score that predicts cognitive decline. The model showed over 80% accuracy and could serve as an early warning system, enabling timely interventions to prevent the progression of MCI and Alzheimer’s disease.

Have you ever wondered why 100 years of research failed so miserably to find dementia treatment? Well, the cause is simple. Almost all treatments are tested in patients living with advanced dementia, that is, in patients who have already lost a significant number of brain cells irreversibly.

This means that if we want to find ways to prevent and manage dementia, we need to find ways to detect it early. It is vital to detect dementia when it starts. It generally starts a decade or even two before its formal diagnosis. Hence, by the time it is diagnosed, an opportunity to reverse dementia in its early stages has been lost.

Fortunately, science has now found one such test, a highly reliable and non-invasive test to predict dementia risk early. This will provide individuals enough time to take measures to prevent overt disease.

Schedule An Appointment With Us

This new study published in Alzheimer’s & Dementia investigated the development of a novel, non-invasive, and highly sensitive EEG-based tool for the early detection of mild cognitive impairment (MCI) among older Black American adults. What is encouraging about this new study or method is that it has an accuracy of above 80%.

The study’s focus on Black seniors was crucial, as they are nearly twice as likely to develop MCI and Alzheimer’s disease (AD) compared to White Americans, yet remained underrepresented in clinical research.

The researchers aimed to create a reliable method that distinguished between healthy controls (HC) and individuals with MCI and provided a “soft score” that predicted the likelihood of cognitive decline.

This approach moved beyond traditional binary classifications by offering an understanding of a participant’s cognitive health. The study used resting-state EEG data, which was more accessible, affordable, and acceptable to patients compared to other neuroimaging techniques.

By analyzing dynamic brain functional connectivity at multiple scales, the researchers developed a model capable of detecting subtle brain activity changes indicative of early cognitive decline.

The study involved 137 Black participants from the Detroit area, including 84 healthy controls and 53 individuals with MCI. EEG data was collected using a high-density cap with 64 active electrodes, and the recordings focused on 12 regions of interest (ROIs) within the brain.

The researchers employed a multiscale sliding window approach to evaluate time-varying functional connectivity between these ROIs. This method allowed them to capture changes in brain activity over time, identifying new biomarkers that reflected the progression from HC to MCI.

To enhance the accuracy and stability of the model, the researchers used a majority voting system among a selected group of classifiers, each based on different window sizes and feature groups. This approach significantly improved the model’s performance, achieving over 91% accuracy in leave-one-out and k-fold cross-validation tests.

The soft discrimination scores generated by the model provided a binary HC or MCI classification and served as indicators of the participants’ overall cognitive health. These scores were shown to predict future cognitive decline, with 84.61% of participants’ progression trends correctly predicted over a 9 to 18-month period.

The study’s findings suggested that this EEG-based tool could serve as an early warning system for cognitive decline, enabling timely interventions that could slow or prevent the progression to MCI and AD.

The soft score approach, in particular, offered a promising avenue for personalized assessments, allowing healthcare providers to identify individuals at high risk for MCI even before clinical symptoms appeared.

In conclusion, this study provided a significant step toward developing a cost-effective, non-invasive, and personalized tool for early detection of cognitive impairment, particularly in underserved populations.

Source:

Deng, J., Sun, B., Kavcic, V., Liu, M., Giordani, B., & Li, T. (2023). Novel methodology for detection and prediction of mild cognitive impairment using resting‐state EEG. Alzheimer’s & Dementia, 20(1), 145–158. https://doi.org/10.1002/alz.13411