國立虎尾科技大學 |

Functional and Effective Connectivity Based Classification and Prediction of Alzheimer's Disease.

紀錄類型:	書目-語言資料,手稿 : Monograph/item
正題名/作者:	Functional and Effective Connectivity Based Classification and Prediction of Alzheimer's Disease./
作者:	Huang, Jiali.
面頁冊數:	1 online resource (115 pages)
附註:	Source: Dissertations Abstracts International, Volume: 84-12, Section: A.
Contained By:	Dissertations Abstracts International84-12A.
標題:	Neuroimaging. -
電子資源:	click for full text (PQDT)
ISBN:	9798379650599

Functional and Effective Connectivity Based Classification and Prediction of Alzheimer's Disease.
Huang, Jiali.

Functional and Effective Connectivity Based Classification and Prediction of Alzheimer's Disease. - 1 online resource (115 pages)

Source: Dissertations Abstracts International, Volume: 84-12, Section: A.

Thesis (Ph.D.)--North Carolina State University, 2023.

Includes bibliographical references

Alzheimer's disease (AD) is one of the most common causes of dementia and it costs hundreds of billions of dollars worldwide each year. There are multiple stages of the disorder: cognitive normal (CN), early mild cognitive impairment (EMCI), late mild cognitive impairment (LMCI), and AD. CN subjects age normally with no sign of depression or dementia, while EMCI and LMCI subjects suffer from difficulties in daily life activity caused by the progressed disease. AD is the advanced and final stage of the disease leading to death. Such a progression can worsen rapidly and cause irreversible changes in the brain without timely detection and proper care. Therefore, accurate understanding and classification of AD are needed as they provide opportunities to initiate proper treatments that may slow the patient's AD progression.The purpose of this dissertation is twofold. First, we study the neural correlates of AD progression. Key regions affected by AD progression were identified using regional homogeneity (ReHo) and amplitude of low-frequency fluctuation (ALFF) measures. Functional connectivity (FC) and effective connectivity (EC) among such regions were calculated. Second, a convolutional neural network (CNN) model was constructed to classify and predict AD progression stages using the connectivity information calculated.Previous studies mainly focused on the structural changes and the time dependencies within the AD brain. Using functional Magnetic Resonance Imaging (fMRI), we analyzed the ReHo and ALFF, representing the regional brain activity levels and the low-frequency oscillations respectively. These measures help highlight several brain regions affected by AD progression and served as a basis for the following connectivity analysis. FC is the temporal correlation between spatially distinct brain regions and is represented by Pearson's correlation. EC represents the causal relationship among regions and is calculated using Dynamic Causal Modeling (DCM) method. Both FC and EC strengths were calculated and we observed that lowered activities in neural correlates are associated with later stages in AD progression.The information derived from FC and EC analyses was then used to train a classifier to distinguish subjects in the four aforementioned AD progression stages. A CNN model was constructed for that purpose. The connectivity matrices were augmented using generative adversarial neural networks (GANs) to prevent overfitting problems related to the limited sample size. Cross-validation was carried out to provide an unbiased evaluation. A 91.95% accuracy rate was achieved, proving the potential of using connectivity to distinguish AD progression stages. The prediction of transitioned subjects did not achieve concise results possibly due to the lack of updated patient documentation and limited input information.By combining FC, EC derived from DCM, and deep learning-based classification, this dissertation revealed the neural correlates, especially causal mechanisms underlying AD progression, and how this facilitates the classification and prediction of AD progression stages. This dissertation laid the groundwork for future connectivity-based neural disorder classification and prediction, which could help in the faster and smarter diagnosis of mental illnesses.

Electronic reproduction.
Ann Arbor, Mich. :
ProQuest,
2024

Mode of access: World Wide Web

ISBN: 9798379650599Subjects--Topical Terms:

1344089
Neuroimaging.
Index Terms--Genre/Form:

554714
Electronic books.

Functional and Effective Connectivity Based Classification and Prediction of Alzheimer's Disease.
LDR:04722ntm a22004097 4500 001 1142787
005 20240429063834.5
006 m o d
007 cr mn ---uuuuu
008 250605s2023 xx obm 000 0 eng d
020 $a 9798379650599
035 $a (MiAaPQ)AAI30463955
035 $a (MiAaPQ)NCState_Univ18402040705
035 $a AAI30463955
040 $a MiAaPQ $b eng $c MiAaPQ $d NTU
100 1 $a Huang, Jiali. $3 1467231
245 1 0 $a Functional and Effective Connectivity Based Classification and Prediction of Alzheimer's Disease.
264 0 $c 2023
300 $a 1 online resource (115 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Dissertations Abstracts International, Volume: 84-12, Section: A.
500 $a Advisor: Fang, Xiaolei;Lalush, David.
502 $a Thesis (Ph.D.)--North Carolina State University, 2023.
504 $a Includes bibliographical references
520 $a Alzheimer's disease (AD) is one of the most common causes of dementia and it costs hundreds of billions of dollars worldwide each year. There are multiple stages of the disorder: cognitive normal (CN), early mild cognitive impairment (EMCI), late mild cognitive impairment (LMCI), and AD. CN subjects age normally with no sign of depression or dementia, while EMCI and LMCI subjects suffer from difficulties in daily life activity caused by the progressed disease. AD is the advanced and final stage of the disease leading to death. Such a progression can worsen rapidly and cause irreversible changes in the brain without timely detection and proper care. Therefore, accurate understanding and classification of AD are needed as they provide opportunities to initiate proper treatments that may slow the patient's AD progression.The purpose of this dissertation is twofold. First, we study the neural correlates of AD progression. Key regions affected by AD progression were identified using regional homogeneity (ReHo) and amplitude of low-frequency fluctuation (ALFF) measures. Functional connectivity (FC) and effective connectivity (EC) among such regions were calculated. Second, a convolutional neural network (CNN) model was constructed to classify and predict AD progression stages using the connectivity information calculated.Previous studies mainly focused on the structural changes and the time dependencies within the AD brain. Using functional Magnetic Resonance Imaging (fMRI), we analyzed the ReHo and ALFF, representing the regional brain activity levels and the low-frequency oscillations respectively. These measures help highlight several brain regions affected by AD progression and served as a basis for the following connectivity analysis. FC is the temporal correlation between spatially distinct brain regions and is represented by Pearson's correlation. EC represents the causal relationship among regions and is calculated using Dynamic Causal Modeling (DCM) method. Both FC and EC strengths were calculated and we observed that lowered activities in neural correlates are associated with later stages in AD progression.The information derived from FC and EC analyses was then used to train a classifier to distinguish subjects in the four aforementioned AD progression stages. A CNN model was constructed for that purpose. The connectivity matrices were augmented using generative adversarial neural networks (GANs) to prevent overfitting problems related to the limited sample size. Cross-validation was carried out to provide an unbiased evaluation. A 91.95% accuracy rate was achieved, proving the potential of using connectivity to distinguish AD progression stages. The prediction of transitioned subjects did not achieve concise results possibly due to the lack of updated patient documentation and limited input information.By combining FC, EC derived from DCM, and deep learning-based classification, this dissertation revealed the neural correlates, especially causal mechanisms underlying AD progression, and how this facilitates the classification and prediction of AD progression stages. This dissertation laid the groundwork for future connectivity-based neural disorder classification and prediction, which could help in the faster and smarter diagnosis of mental illnesses.
533 $a Electronic reproduction. $b Ann Arbor, Mich. : $c ProQuest, $d 2024
538 $a Mode of access: World Wide Web
650 4 $a Neuroimaging. $3 1344089
650 4 $a Deep learning. $3 1127428
650 4 $a Memory. $3 555749
650 4 $a Attention. $3 555203
650 4 $a Brain research. $3 1372495
650 4 $a Electroencephalography. $3 581307
650 4 $a Cognitive ability. $3 1372497
650 4 $a Medical imaging. $3 1180167
650 4 $a Cognition & reasoning. $3 1372461
650 4 $a Alzheimers disease. $3 1413648
650 4 $a Neurological disorders. $3 1467232
650 4 $a Magnetic resonance imaging. $3 644374
650 4 $a Dementia. $3 678297
650 4 $a Neural networks. $3 1011215
650 4 $a Autism. $3 583598
650 4 $a Biomarkers. $3 1155103
650 4 $a Aging. $3 559847
650 4 $a Cognitive psychology. $3 556029
650 4 $a Disability studies. $3 809395
650 4 $a Gerontology. $3 559846
650 4 $a Neurosciences. $3 593561
650 4 $a Psychology. $3 555998
655 7 $a Electronic books. $2 local $3 554714
690 $a 0800
690 $a 0574
690 $a 0493
690 $a 0633
690 $a 0201
690 $a 0351
690 $a 0317
690 $a 0621
710 2 $a ProQuest Information and Learning Co. $3 1178819
710 2 $a North Carolina State University. $3 845424
773 0 $t Dissertations Abstracts International $g 84-12A.
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=30463955 $z click for full text (PQDT)