Alzheimer’s diseases diagnosis using fusion of high informative BiLSTM and CNN features of EEG signal

Biomedical Signal Processing and Control, Journal Year: 2023, Volume and Issue: 86, P. 105298 - 105298

Published: July 29, 2023

Language: Английский

---

Alzheimer’s disease diagnosis from single and multimodal data using machine and deep learning models: Achievements and future directions

Expert Systems with Applications, Journal Year: 2024, Volume and Issue: 255, P. 124780 - 124780

Published: July 14, 2024

Language: Английский

---

Brain health in diverse settings: How age, demographics and cognition shape brain function

NeuroImage, Journal Year: 2024, Volume and Issue: 295, P. 120636 - 120636

Published: May 21, 2024

Diversity in brain health is influenced by individual differences demographics and cognition. However, most studies on diseases have typically controlled for these factors rather than explored their potential to predict signals. Here, we assessed the role of (age, sex, education; n = 1,298) cognition (n 725) as predictors different metrics usually used case-control studies. These included power spectrum aperiodic (1/f slope, knee, offset) metrics, well complexity (fractal dimension estimation, permutation entropy, Wiener spectral structure variability) connectivity (graph-theoretic mutual information, conditional organizational information) from source space resting-state EEG activity a diverse sample global south north populations. Brain-phenotype models were computed using reflecting local (power components) dynamics interactions (complexity graph-theoretic measures). Electrophysiological modulated despite varied methods data acquisition assessments across multiple centers, indicating that results unlikely be accounted methodological discrepancies. Variations signals mainly age cognition, while education sex exhibited less importance. Power measures sensitive capturing differences. Older age, poorer being male associated with reduced alpha power, whereas older network integration segregation. Findings suggest basic impact core function are standard Considering variability diversity settings would contribute more tailored understanding function.

Language: Английский

---

A Comparative Study on Feature Extraction Techniques for the Discrimination of Frontotemporal Dementia and Alzheimer’s Disease with Electroencephalography in Resting-State Adults

Brain Sciences, Journal Year: 2024, Volume and Issue: 14(4), P. 335 - 335

Published: March 29, 2024

Early-stage Alzheimer’s disease (AD) and frontotemporal dementia (FTD) share similar symptoms, complicating their diagnosis the development of specific treatment strategies. Our study evaluated multiple feature extraction techniques for identifying AD FTD biomarkers from electroencephalographic (EEG) signals. We developed an optimised machine learning architecture that integrates sliding windowing, extraction, supervised to distinguish between patients, as well healthy controls (HCs). model, with a 90% overlap SVD entropy K-Nearest Neighbors (KNN) learning, achieved mean F1-score accuracy 93% 91%, 92.5% 93%, 91.5% 91% discriminating HC, FTD, respectively. The importance array, explainable AI feature, highlighted brain lobes contributed distinguishing biomarkers. This research introduces novel framework detecting using EEG signals, addressing need accurate early-stage diagnostics. Furthermore, comparative evaluation methods on AD/FTD detection discrimination is documented.

Language: Английский

---

DCNN-SBiL: EEG signal based mild cognitive impairment classification using compact convolutional network

Expert Systems with Applications, Journal Year: 2025, Volume and Issue: unknown, P. 126553 - 126553

Published: Jan. 1, 2025

Language: Английский

---

An explainable Artificial Intelligence approach to study MCI to AD conversion via HD-EEG processing

Clinical EEG and Neuroscience, Journal Year: 2021, Volume and Issue: 54(1), P. 51 - 60

Published: Dec. 10, 2021

An explainable Artificial Intelligence (xAI) approach is proposed to longitudinally monitor subjects affected by Mild Cognitive Impairment (MCI) using high-density electroencephalography (HD-EEG). To this end, a group of MCI patients was enrolled at IRCCS Centro Neurolesi Bonino Pulejo Messina (Italy) within follow-up protocol that included two evaluations steps: T0 (first evaluation) and T1 (three months later). At T1, four converted Alzheimer’s Disease (AD) were in the analysis as goal work use xAI detect individual changes EEGs possibly related degeneration from AD. The methodology consists mapping segments HD-EEG into channel-frequency maps means power spectral density. Such are used input Convolutional Neural Network (CNN), trained label “T0” (MCI state) or “T1” (AD state). Experimental results reported high intra-subject classification performance (accuracy rate up 98.97% (95% confidence interval: 98.68–99.26)). Subsequently, explainability CNN explored via Grad-CAM approach. procedure detected which EEG-channels (i.e., head region) range frequencies sub-bands) more active progression showed main information delta sub-band that, limited analyzed dataset, highest relevant areas are: left-temporal central-frontal lobe for Sb01, parietal Sb02, left-frontal Sb03 left-frontotemporal region Sb04.

Language: Английский

---

20 years of ordinal patterns: Perspectives and challenges

EPL (Europhysics Letters), Journal Year: 2022, Volume and Issue: 138(3), P. 31001 - 31001

Published: April 26, 2022

In 2002, in a seminal article, Christoph Bandt and Bernd Pompe proposed new methodology for the analysis of complex time series, now known as Ordinal Analysis. The ordinal is based on computation symbols (known patterns) which are defined terms temporal ordering data points whose probabilities probabilities. With probabilities, Shannon entropy can be calculated, permutation entropy. Since it was proposed, method has found applications fields diverse biomedicine climatology. However, some properties still not fully understood, how to combine approach feature extraction with machine learning techniques model identification, series classification or forecasting remains challenge. objective this perspective article present recent advances discuss open problems.

Language: Английский

---

A self-driven approach for multi-class discrimination in Alzheimer's disease based on wearable EEG

Computer Methods and Programs in Biomedicine, Journal Year: 2022, Volume and Issue: 220, P. 106841 - 106841

Published: April 26, 2022

Early detection is critical to control Alzheimer's disease (AD) progression and postpone cognitive decline. Traditional medical procedures such as magnetic resonance imaging are costly, involve long waiting lists, require complex analysis. Alternatively, for the past years, researchers have successfully evaluated AD approaches based on machine learning electroencephalography (EEG). Nonetheless, these frequently rely upon manual processing or non-portable EEG hardware. These aspects suboptimal regarding automated diagnosis, since they additional personnel hinder portability. In this work, we report preliminary evaluation of a self-driven multi-class discrimination approach commercial acquisition system using sixteen channels. For purpose, recorded three groups participants: mild AD, impairment (MCI) non-AD, controls, implemented analysis pipeline discriminate groups. First, applied artifact rejection algorithms recordings. Then, extracted power, entropy, complexity features from preprocessed epochs. Finally, classification problem multi-layer perceptron through leave-one-subject-out cross-validation. The results that obtained comparable best in literature (0.88 F1-score), what suggests can potentially be detected learning. We believe work further research could contribute opening door single consultation session, therefore reducing costs associated screening advancing treatment.

Language: Английский

---

A systematic review and methodological analysis of EEG-based biomarkers of Alzheimer's disease

Measurement, Journal Year: 2023, Volume and Issue: 220, P. 113274 - 113274

Published: July 4, 2023

Language: Английский

---

Diagnosis of Alzheimer’s disease via resting-state EEG: integration of spectrum, complexity, and synchronization signal features

Frontiers in Aging Neuroscience, Journal Year: 2023, Volume and Issue: 15

Published: Nov. 7, 2023

Alzheimer's disease (AD) is the most common neurogenerative disorder, making up 70% of total dementia cases with a prevalence more than 55 million people. Electroencephalogram (EEG) has become suitable, accurate, and highly sensitive biomarker for identification diagnosis AD.In this study, public database EEG resting state-closed eye recordings containing 36 AD subjects 29 normal was used. And then, three types signal features resting-state EEG, i.e., spectrum, complexity, synchronization, were performed by applying various processing statistical methods, to obtain 18 each epoch. Next, supervised machine learning classification algorithms decision trees, random forests, support vector (SVM) compared in categorizing processed leave-one-person-out cross-validation.The results showed that cases, major change characteristics an slowing, reduced decrease synchrony. The proposed methodology achieved relatively high accuracy 95.65, 95.86, 88.54% between SVM, respectively, showing integration synchronization signals can enhance performance identifying subjects.This study recommended aiding AD.

Language: Английский

---