Subject classification and cross-time prediction based on functional connectivity and white matter microstructure features in a rat model of Alzheimer’s using machine learning DOI Creative Commons
Yujian Diao, Ileana Jelescu

bioRxiv (Cold Spring Harbor Laboratory), Год журнала: 2023, Номер unknown

Опубликована: Март 27, 2023

Abstract Background The pathological process of Alzheimer’s disease (AD) typically takes up decades from onset to clinical symptoms. Early brain changes in AD include MRI-measurable features such as aItered functional connectivity (FC) and white matter degeneration. ability these discriminate between subjects without a diagnosis, or their prognostic value, is however not established. Methods main trigger mechanism still debated, although impaired glucose metabolism taking an increasingly central role. Here we used rat model sporadic AD, based on induced by intracerebroventricular injection streptozotocin (STZ). We characterized alterations FC microstructure longitudinally using diffusion MRI. Those MRI-derived measures were classify STZ control rats machine learning, the importance each individual measure was quantified explainable artificial intelligence methods. Results Overall, combining all metrics ensemble way best strategy rats, with consistent accuracy over 0.85. However, early achieved features, later FC. This suggests that damage group might precede For cross-timepoint prediction, also had highest performance while, contrast, reduced its dynamic pattern which shifted hyperconnectivity late hypoconnectivity. Conclusions Our study highlights vs course disease, potential translation humans.

Язык: Английский

Computational approaches to Explainable Artificial Intelligence: Advances in theory, applications and trends DOI Creative Commons
J. M. Górriz, I. Álvarez, Agustín Álvarez-Marquina

и другие.

Information Fusion, Год журнала: 2023, Номер 100, С. 101945 - 101945

Опубликована: Июль 29, 2023

Deep Learning (DL), a groundbreaking branch of Machine (ML), has emerged as driving force in both theoretical and applied Artificial Intelligence (AI). DL algorithms, rooted complex non-linear artificial neural systems, excel at extracting high-level features from data. demonstrated human-level performance real-world tasks, including clinical diagnostics, unlocked solutions to previously intractable problems virtual agent design, robotics, genomics, neuroimaging, computer vision, industrial automation. In this paper, the most relevant advances last few years (AI) several applications neuroscience, robotics are presented, reviewed discussed. way, we summarize state-of-the-art AI methods, models within collection works presented 9th International Conference on Interplay between Natural Computation (IWINAC). The paper excellent examples new scientific discoveries made laboratories that have successfully transitioned real-life applications.

Язык: Английский

Процитировано

104
A Meta-Heuristic Multi-Objective Optimization Method for Alzheimer’s Disease Detection Based on Multi-Modal Data DOI Creative Commons
Walaa N. Ismail,

Fathimathul Rajeena P. P.,

Mona A. S. Ali

и другие.

Mathematics, Год журнала: 2023, Номер 11(4), С. 957 - 957

Опубликована: Фев. 13, 2023

Alzheimer’s disease (AD) is a neurodegenerative that affects large number of people across the globe. Even though AD one most commonly seen brain disorders, it difficult to detect and requires categorical representation features differentiate similar patterns. Research into more complex problems, such as detection, frequently employs neural networks. Those approaches are regarded well-understood even sufficient by researchers scientists without formal training in artificial intelligence. Thus, imperative identify method detection fully automated user-friendly non-AI experts. The should find efficient values for models’ design parameters promptly simplify network process subsequently democratize Further, multi-modal medical image fusion has richer modal superior ability represent information. A formed integrating relevant complementary information from multiple input images facilitate accurate diagnosis better treatment. This study presents MultiAz-Net novel optimized ensemble-based deep learning model incorporate heterogeneous PET MRI diagnose disease. Based on extracted fused data, we propose an procedure predicting onset at early stage. Three steps involved proposed architecture: fusion, feature extraction, classification. Additionally, Multi-Objective Grasshopper Optimization Algorithm (MOGOA) presented multi-objective optimization algorithm optimize layers MultiAz-Net. desired objective functions imposed achieve this, searched corresponding values. ensemble been tested perform four categorization tasks, three binary categorizations, multi-class task utilizing publicly available Alzheimer neuroimaging dataset. achieved (92.3 ± 5.45)% accuracy multi-class-classification task, significantly than other models have reported.

Язык: Английский

Процитировано

32
Increased functional connectivity patterns in mild Alzheimer’s disease: A rsfMRI study DOI Creative Commons
Lucía Penalba‐Sánchez, Patrícia Oliveira‐Silva, Alexander Sumich

и другие.

Frontiers in Aging Neuroscience, Год журнала: 2023, Номер 14

Опубликована: Янв. 9, 2023

Alzheimer's disease (AD) is the most common age-related neurodegenerative disorder. In view of our rapidly aging population, there an urgent need to identify at early stage. A potential way do so by assessing functional connectivity (FC), i.e., statistical dependency between two or more brain regions, through novel analysis techniques.In present study, we assessed static and dynamic FC using different approaches. resting state (rs)fMRI dataset from neuroimaging initiative (ADNI) was used (n = 128). The blood-oxygen-level-dependent (BOLD) signals 116 regions 4 groups participants, healthy controls (HC; n 35), mild cognitive impairment (EMCI; 29), late (LMCI; 30), (AD; 34) were extracted analyzed. Pearson's correlation, sliding-windows correlation (SWA), point process (PPA). Additionally, graph theory measures explore network segregation integration computed.Our results showed a longer characteristic path length decreased degree EMCI in comparison other groups. increased several LMCI AD contrast HC detected. These suggest maladaptive short-term mechanism maintain cognition.The pattern observable all analyses; however, PPA enabled us reduce computational demands offered new specific findings.

Язык: Английский

Процитировано

25
Atrophy of the cholinergic regions advances from early to late mild cognitive impairment DOI

Ying-Liang Larry Lai,

Fei‐Ting Hsu,

Shu-Yi Yeh

и другие.

Neuroradiology, Год журнала: 2024, Номер 66(4), С. 543 - 556

Опубликована: Янв. 19, 2024

Язык: Английский

Процитировано

3
Comparison of the diagnostic accuracy of resting-state fMRI driven machine learning algorithms in the detection of mild cognitive impairment DOI Creative Commons

Gergo Bolla,

Dalida Borbála Berente,

Anita Andrássy

и другие.

Scientific Reports, Год журнала: 2023, Номер 13(1)

Опубликована: Дек. 14, 2023

Abstract Mild cognitive impairment (MCI) is a potential therapeutic window in the prevention of dementia; however, automated detection early deterioration an unresolved issue. The aim our study was to compare various classification approaches differentiate MCI patients from healthy controls, based on rs-fMRI data, using machine learning (ML) algorithms. Own dataset (from two centers) and ADNI database were used during analysis. Three fMRI parameters applied five feature selection algorithms: local correlation, intrinsic connectivity, fractional amplitude low frequency fluctuations. Support vector (SVM) random forest (RF) methods for classification. We achieved relatively wide range 78–87% accuracy with SVM combining three parameters. In datasets case we can also see even 90% scores. RF provided more harmonized result among algorithms both 80–84% 74–82% database. Despite some lower performance metrics algorithms, most results positive could be seen unrelated which increase validity methods. Our highlight ML-based applications diagnostic techniques recognize patients.

Язык: Английский

Процитировано

7
Alteration in amygdala subfield volumes and their association with cognition in mild cognitive impairment DOI
Sadhana Singh, Palash Kumar Malo, Albert Stezin

и другие.

Journal of Neurology, Год журнала: 2024, Номер 271(8), С. 5460 - 5467

Опубликована: Июнь 15, 2024

Язык: Английский

Процитировано

2
Multivariate pattern analysis of medical imaging-based Alzheimer's disease DOI Creative Commons
Maitha Alarjani, Badar Almarri

Frontiers in Medicine, Год журнала: 2024, Номер 11

Опубликована: Июль 19, 2024

Alzheimer's disease (AD) is a devastating brain disorder that steadily worsens over time. It marked by relentless decline in memory and cognitive abilities. As the progresses, it leads to significant loss of mental function. Early detection AD essential starting treatments can mitigate progression this enhance patients' quality life. This study aims observe AD's functional connectivity pattern extract patterns through multivariate analysis (MVPA) analyze activity across multiple voxels. The optimized feature extraction techniques are used obtain important features for performing training on models using several hybrid machine learning classifiers binary classification multi-class classification. proposed approach has been applied two public datasets named Open Access Series Imaging Studies (OASIS) Neuroimaging Initiative (ADNI). results evaluated performance metrics, comparisons have made differentiate between different stages visualization tools.

Язык: Английский

Процитировано

2
Functional Brain Network Measures for Alzheimer’s Disease Classification DOI Creative Commons
Luyun Wang, Jinhua Sheng, Qiao Zhang

и другие.

IEEE Access, Год журнала: 2023, Номер 11, С. 111832 - 111845

Опубликована: Янв. 1, 2023

Background: Alzheimer's disease (AD) is an incurable neurodegenerative primarily affecting the elderly population. The therapy of AD depends heavily on early diagnosis. In this study, our primary objective to evaluate classification framework, which combines graph theory and machine learning techniques for functional magnetic resonance imaging (fMRI), distinguish AD, mild cognitive impairment (EMCI), late (LMCI), healthy control (HC). Methods: A novel multi-feature selection method, incorporating dual theoretical approach, proposed classification. This method utilizes three different feature methods after brain areas through graph-theory analyses in 96 subjects with parcellation by using joint human connectome project multimodal (J-HCPMMP) 180 per hemisphere. Results: results show that optimal features selected minimal redundancy maximal relevance (MRMR) based support vector linear (SVM-linear) from measures 36 360 areas. accuracies identifying HC vs. EMCI, LMCI, EMCI LMCI are 85.60%, 92.90%, 96.80%, 83.30%, 84.90% 89.50%, respectively. Conclusion: indicate combination fMRI connectivity analysis might be helpful diagnosis especially use local measures, may better reflect changes regions because impairment.

Язык: Английский

Процитировано

5
Diagnosis of early mild cognitive impairment using a multiobjective optimization algorithm based on T1-MRI data DOI Creative Commons
Jafar Zamani, Ali Sadr, Amir‐Homayoun Javadi

и другие.

Scientific Reports, Год журнала: 2022, Номер 12(1)

Опубликована: Янв. 19, 2022

Abstract Alzheimer’s disease (AD) is the most prevalent form of dementia. The accurate diagnosis AD, especially in early phases very important for timely intervention. It has been suggested that brain atrophy, as measured with structural magnetic resonance imaging (sMRI), can be an efficacy marker neurodegeneration. While classification methods have successful performance such poor those stages mild cognitive impairment (EMCI). Therefore, this study we investigated whether optimisation based on evolutionary algorithms (EA) effective tool EMCI compared to cognitively normal participants (CNs). Structural MRI data patients (n = 54) and CN 56) was extracted from Neuroimaging Initiative (ADNI). Using three automatic segmentation methods, volumetric parameters input algorithms. Our method achieved accuracy greater than 93%. This level higher previously using a single- or multiple modalities data. results show method, single modality biomarkers enough achieve high accuracy.

Язык: Английский

Процитировано

8
fMRI-based Alzheimer’s disease detection via functional connectivity analysis: a systematic review DOI Creative Commons
Maitha Alarjani, Badar Almarri

PeerJ Computer Science, Год журнала: 2024, Номер 10, С. e2302 - e2302

Опубликована: Окт. 16, 2024

Alzheimer’s disease is a common brain disorder affecting many people worldwide. It the primary cause of dementia and memory loss. The early diagnosis essential to provide timely care AD patients prevent development symptoms this disease. Various non-invasive techniques can be utilized diagnose in its stages. These include functional magnetic resonance imaging, electroencephalography, positron emission tomography, diffusion tensor imaging. They are mainly used explore structural connectivity human brains. Functional for understanding co-activation certain regions co-activation. This systematic review scrutinizes various works detection by analyzing learning from fMRI datasets that were published between 2018 2024. work investigates whole pipeline including data analysis, standard preprocessing phases fMRI, feature computation, extraction selection, machine deep algorithms predict occurrence Ultimately, paper analyzed results on highlighted future research directions medical There need an efficient accurate way detect overcome problems faced

Язык: Английский

Процитировано

1