Identifying the mechanism to forecast the progression of Alzheimer’s disease from mild cognitive impairment using deep learning
Alzheimer’s disease is a progressive neurodegenerative disease affecting the cognitive and behavioural functions. Although there is currently no cure for AD, early diagnosis is crucial for doctors to intervene and decide on management plans. In the recent years, computer aided diagnosis method has b...
| Main Author: | |
|---|---|
| Format: | Final Year Project / Dissertation / Thesis |
| Published: |
2022
|
| Subjects: | |
| Online Access: | http://eprints.utar.edu.my/5231/ http://eprints.utar.edu.my/5231/1/BI_1900392_Final_%2D_HIU_WEI_XIN_THERESA.pdf |
| _version_ | 1848886361758629888 |
|---|---|
| author | Hiu, Theresa Wei Xin |
| author_facet | Hiu, Theresa Wei Xin |
| author_sort | Hiu, Theresa Wei Xin |
| building | UTAR Institutional Repository |
| collection | Online Access |
| description | Alzheimer’s disease is a progressive neurodegenerative disease affecting the cognitive and behavioural functions. Although there is currently no cure for AD, early diagnosis is crucial for doctors to intervene and decide on management plans. In the recent years, computer aided diagnosis method has been employed for the detection and diagnosis of the progression of AD. Thus, this study aims to identify the progression of AD for patients with MCI using only structural MRI data. This study devised a structured method in classifying the subjects into CN, AD, sMCI and pMCI groups. This study also implemented a CNN algorithm based on 3D ResNet-18 model using weights from ImageNet for the classification task of CN vs AD and sMCI vs pMCI. Visualization of the CNN decision using Grad-CAM was used for the detection of most important biomarkers and identification of the discriminative brain regions related to AD and pMCI. The results demonstrated by the CNN model for the CN vs AD classification task achieved an accuracy of 75% and AUC of 0.81. The CNN model used to predict the conversion of MCI patients to AD (sMCI vs pMCI) achieved an accuracy of 65.91% and AUC of 0.60. Based on the Grad-CAM visualization, the regions of the brain surrounding the hippocampus and amygdala was found to have contributed to the CNN prediction. CNN models and visualization of the brain regions which contributed to the CNN prediction assist doctors in the diagnosis of MCI conversion and the discovery of new biomarkers. |
| first_indexed | 2025-11-15T19:37:17Z |
| format | Final Year Project / Dissertation / Thesis |
| id | utar-5231 |
| institution | Universiti Tunku Abdul Rahman |
| institution_category | Local University |
| last_indexed | 2025-11-15T19:37:17Z |
| publishDate | 2022 |
| recordtype | eprints |
| repository_type | Digital Repository |
| spelling | utar-52312023-03-08T07:59:11Z Identifying the mechanism to forecast the progression of Alzheimer’s disease from mild cognitive impairment using deep learning Hiu, Theresa Wei Xin R Medicine (General) TA Engineering (General). Civil engineering (General) Alzheimer’s disease is a progressive neurodegenerative disease affecting the cognitive and behavioural functions. Although there is currently no cure for AD, early diagnosis is crucial for doctors to intervene and decide on management plans. In the recent years, computer aided diagnosis method has been employed for the detection and diagnosis of the progression of AD. Thus, this study aims to identify the progression of AD for patients with MCI using only structural MRI data. This study devised a structured method in classifying the subjects into CN, AD, sMCI and pMCI groups. This study also implemented a CNN algorithm based on 3D ResNet-18 model using weights from ImageNet for the classification task of CN vs AD and sMCI vs pMCI. Visualization of the CNN decision using Grad-CAM was used for the detection of most important biomarkers and identification of the discriminative brain regions related to AD and pMCI. The results demonstrated by the CNN model for the CN vs AD classification task achieved an accuracy of 75% and AUC of 0.81. The CNN model used to predict the conversion of MCI patients to AD (sMCI vs pMCI) achieved an accuracy of 65.91% and AUC of 0.60. Based on the Grad-CAM visualization, the regions of the brain surrounding the hippocampus and amygdala was found to have contributed to the CNN prediction. CNN models and visualization of the brain regions which contributed to the CNN prediction assist doctors in the diagnosis of MCI conversion and the discovery of new biomarkers. 2022 Final Year Project / Dissertation / Thesis NonPeerReviewed application/pdf http://eprints.utar.edu.my/5231/1/BI_1900392_Final_%2D_HIU_WEI_XIN_THERESA.pdf Hiu, Theresa Wei Xin (2022) Identifying the mechanism to forecast the progression of Alzheimer’s disease from mild cognitive impairment using deep learning. Final Year Project, UTAR. http://eprints.utar.edu.my/5231/ |
| spellingShingle | R Medicine (General) TA Engineering (General). Civil engineering (General) Hiu, Theresa Wei Xin Identifying the mechanism to forecast the progression of Alzheimer’s disease from mild cognitive impairment using deep learning |
| title | Identifying the mechanism to forecast the progression of Alzheimer’s disease from mild cognitive impairment using deep learning |
| title_full | Identifying the mechanism to forecast the progression of Alzheimer’s disease from mild cognitive impairment using deep learning |
| title_fullStr | Identifying the mechanism to forecast the progression of Alzheimer’s disease from mild cognitive impairment using deep learning |
| title_full_unstemmed | Identifying the mechanism to forecast the progression of Alzheimer’s disease from mild cognitive impairment using deep learning |
| title_short | Identifying the mechanism to forecast the progression of Alzheimer’s disease from mild cognitive impairment using deep learning |
| title_sort | identifying the mechanism to forecast the progression of alzheimer’s disease from mild cognitive impairment using deep learning |
| topic | R Medicine (General) TA Engineering (General). Civil engineering (General) |
| url | http://eprints.utar.edu.my/5231/ http://eprints.utar.edu.my/5231/1/BI_1900392_Final_%2D_HIU_WEI_XIN_THERESA.pdf |