Graph Convolutional Networks For Disease Mapping and Classification in Healthcare DOI
Rakesh Kumar, Devvret Verma,

J. Relin Francis Raj

et al.

Published: Dec. 29, 2023

In the context of healthcare, this study investigates use Graph A convolutional Networks (GCNs) for disease mapping along with classification. Based on an interpretivist philosophical thought, a descriptive design alongside secondary data collection is used in deductive manner. The research creates strong framework sickness mapping, assesses how well GCNs adapt to varied health information, and compares their effectiveness more conventional machine learning techniques order determine suitable they are. An investigation conducted into understanding GCN-based diagnosis models, offering valuable perspectives decision-making procedures. findings support improved diagnostic precision, wellinformed treatment planning, precision medical treatments. emphasis when applying results procedures connection systems that provide decision support, ongoing improvement. importance model interpretability, ability be general as realworld integration highlighted by critical analysis. Developing interpretability strategies addressing ethical issues are among recommendations. ensure responsible deployment, future work ought concentrate improving GCN architectures, integrating multi-modal information advocating interdisciplinary collaboration.

Language: Английский

Extensive T1-weighted MRI preprocessing improves generalizability of deep brain age prediction models DOI Creative Commons
Lara Dular,

Franjo Pernuš,

Žiga Špiclin

et al.

Computers in Biology and Medicine, Journal Year: 2024, Volume and Issue: 173, P. 108320 - 108320

Published: March 20, 2024

Brain age is an estimate of chronological obtained from T1-weighted magnetic resonance images (T1w MRI), representing a straightforward diagnostic biomarker brain aging and associated diseases. While the current best accuracy predictions on T1w MRIs healthy subjects ranges two to three years, comparing results across studies challenging due differences in datasets, preprocessing pipelines, evaluation protocols used. This paper investigates impact image performance four deep learning models recent literature. Four which differed terms registration transform, grayscale correction, software implementation, were evaluated. The showed that choice or steps could significantly affect prediction error, with maximum increase 0.75 years mean absolute error (MAE) for same model dataset. correction had no significant MAE, using affine rather than rigid atlas statistically improved MAE. Models trained 3D isotropic 1mm3 resolution exhibited less sensitivity variations compared 2D those downsampled images. Our findings indicate extensive improves especially when predicting new runs counter prevailing research literature, suggests minimally preprocessed scans are better suited unseen scanners. We demonstrate that, irrespective used during training, applying some form offset essential enable model's generalize effectively datasets sites, regardless whether they have undergone different as training set.

Language: Английский

Citations

9
eXplainable Artificial Intelligence (XAI) in aging clock models DOI
Alena Kalyakulina, Igor Yusipov, Alexey Moskalev

et al.

Ageing Research Reviews, Journal Year: 2023, Volume and Issue: 93, P. 102144 - 102144

Published: Nov. 28, 2023

Language: Английский

Citations

19
A Review of Graph Theory-Based Diagnosis of Neurological Disorders Based on EEG and MRI DOI
Ying Yan, Guanting Liu, Haoyang Cai

et al.

Neurocomputing, Journal Year: 2024, Volume and Issue: 599, P. 128098 - 128098

Published: Sept. 1, 2024

Language: Английский

Citations

6
LSTGINet: Local Attention Spatio-Temporal Graph Inference Network for Age Prediction DOI Creative Commons

Yi Lei,

Xin Wen, Yanrong Hao

et al.

Algorithms, Journal Year: 2025, Volume and Issue: 18(3), P. 138 - 138

Published: March 3, 2025

There is a close correlation between brain aging and age. However, traditional neural networks cannot fully capture the potential age due to limited receptive field. Furthermore, they are more concerned with deep spatial semantics, ignoring fact that effective temporal information can enrich representation of low-level semantics. To address these limitations, local attention spatio-temporal graph inference network (LSTGINet) was developed explore details association aging, taking into account both perspectives. First, multi-scale branches used increase field model simultaneously, achieving perception static correlation. Second, feature graphs reconstructed, large topographies constructed. The node aggregation transfer functions hidden dynamic A new module embedded in component global context establish dependencies interactivity different features, balance differences distribution We use newly designed weighted loss function supervise learning entire prediction framework strengthen process final experimental results show MAE on baseline datasets such as CamCAN NKI 6.33 6.28, respectively, better than current state-of-the-art methods, provides basis for assessing state adults.

Language: Английский

Citations

0
Evaluating the effects of volume censoring on fetal functional connectivity DOI Creative Commons

Jung‐Hoon Kim,

Josepheen De Asis‐Cruz, Kevin M. Cook

et al.

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: April 16, 2025

Abstract Advances in neuroimaging have enabled non-invasive investigation of fetal brain development vivo. Resting-state functional magnetic resonance imaging (rs-fMRI) has provided critical insights into emerging networks fetuses. However, acquiring high-quality rs-fMRI remains challenging due to the unpredictable and unconstrained motion head. Nuisance regression, where signal is regressed onto translational rotational head parameters, been widely effectively used adults reduce influence motion. subsequent studies revealed that associations between large-scale connectivity (FC) persisted even after regression. In ex utero groups (e.g., newborns, toddlers, adults), censoring high-motion volumes shown effectiveness mitigating such lingering impacts While high utilized rs-fMRI, a systematic assessment regression fetuses not done. Establishing avoid possible bias findings resulting from To address this knowledge gap, we investigated at different analysis scales: blood oxygenation level dependent (BOLD) time series whole-brain FC. We dataset 120 scans collected 104 healthy found nuisance reduced association motion, defined by frame-by-frame displacement (FD) position, BOLD data all regions interest (ROI) encompassing whole brain. however, was effective reducing impact on Fetuses’ FC profiles significantly predicted average FD ( r = 0.09 ± 0.08; p < 10 –3 ) suggesting effect patterns. dissociate FC, volume evaluated its efficacy correcting thresholds. demonstrated censored improved resting state data’s ability predict neurobiological features, as gestational age sex (accuracy 55.2 2.9% with 1.5 mm vs. 44.6 3.6% no censoring). Collectively, our results highlight importance thus attenuating motion-related bias. Like older neonates adults, combining techniques recommended for analysis, e.g., network-based

Language: Английский

Citations

0
Investigating the relationship between brain age and Alzheimer’s disease: A deep learning approach with multimodal MRI DOI
Zhengning Wang, Jiaxin Liu, Fang Chen

et al.

Biomedical Signal Processing and Control, Journal Year: 2025, Volume and Issue: 109, P. 107926 - 107926

Published: May 6, 2025

Language: Английский

Citations

0
Brain age prediction from MRI images based on a convolutional neural network with MRMR feature selection layer DOI Creative Commons

Mustafa Hatem Al Ghariri,

Seyed Omid Shahdi

Wasit Journal of Computer and Mathematics Science, Journal Year: 2025, Volume and Issue: 4(1), P. 17 - 29

Published: March 30, 2025

An sophisticated medical technique used to diagnose illnesses and brain disorders including multiple sclerosis, Alzheimer's, other neurological ailments is the ability predict biological age of using MRI pictures. To do this, algorithms neural networks are scan pictures in order extract different properties, cortical thickness volume. The ages individuals determined by matching their characteristics against imaging data collected from patients. research employs a new deep learning model named CNN-MRMR which combines features Minimum Redundancy Maximum Relevance (MRMR) feature selection approach Convolutional Neural Network (CNN) technology. images human brains initially processed convolutional network age-related characteristics. layer uses MRMR algorithm identifies essential for target variable while minimizing redundancy select optimal subset. system regression as final stage utilizing selected proposed method estimating individual attained prediction accuracy 90.3%, outperforming results comparable studies.

Language: Английский

Citations

0
Echo-GRU: Emotion Recognition Using Wearable EEG Supporting Early Alzheimer’s Disease Detection DOI
Quoc-Toan Nguyen

Lecture notes in computer science, Journal Year: 2025, Volume and Issue: unknown, P. 3 - 17

Published: Jan. 1, 2025

Language: Английский

Citations

0
Recent Developments in Neuroinformatics and Computational Neuroscience DOI Creative Commons

Polly Jones

Journal of Biomedical and Sustainable Healthcare Applications, Journal Year: 2023, Volume and Issue: unknown, P. 118 - 128

Published: July 5, 2023

In comparison to other natural systems, the temporal dynamics of human brain's growth, structure, and function are notably intricate. The brain is comprised an estimated 86.1 8.0 billion neurons a comparable non-neural glial cells number. Additionally, contains neuronal systems with over 100 trillion connections. modeling, analysis, comprehension these complex structures require use code automation. Neuroinformatics methodologies employed manage, retrieve, integrate copious quantities data produced through clinical documentation, scientific literature, specialized databases. Conversely, computational neuroscience, which draws heavily upon fields biology, physics, mathematics, computation, tackles issues. interdisciplinary field that integrates neuroscience neuroscientific experimentation. This paper functions as introductory guide for individuals who lack familiarity domains neuroinformatics along their consistentsophisticated software, resources, tools.

Language: Английский

Citations

7
Brain age as a biomarker for pathological versus healthy ageing – a REMEMBER study DOI Creative Commons
Mandy Melissa Jane Wittens, Stijn Denissen, Diana M. Sima

et al.

Alzheimer s Research & Therapy, Journal Year: 2024, Volume and Issue: 16(1)

Published: June 14, 2024

This study aimed to evaluate the potential clinical value of a new brain age prediction model as single interpretable variable representing condition our brain. Among many use cases, could be novel outcome measure assess preventive effect life-style interventions. The REMEMBER population (N = 742) consisted cognitively healthy (HC,N 91), subjective cognitive decline (SCD,N 65), mild impairment (MCI,N 319) and AD dementia (ADD,N 267) subjects. Automated volumetry global, cortical, subcortical structures computed by CE-labeled FDA-cleared software icobrain dm (dementia) was retrospectively extracted from T1-weighted MRI sequences that were acquired during routine at participating memory clinics Belgian Dementia Council. volumetric features, along with sex, combined into weighted sum using linear model, used predict 'brain age' predicted difference' (BPAD age-chronological age) for every subject. MCI ADD patients showed an increased compared their chronological age. Overall, outperformed BPAD in terms classification accuracy across spectrum. There weak-to-moderate correlation between total MMSE score both (r -0.38,p < .001) -0.26,p .001). Noticeable trends, but no significant correlations, found incidence conversion ADD, nor time ADD. heavy alcohol drinkers non-/sporadic (p .014) moderate .040) drinkers. Brain associated have serve indicators for, impact lifestyle modifications or interventions on, health.

Language: Английский

Citations

2