There are many active research projects accessing and applying shared ADNI data. Use the search above to find specific research focuses on the active ADNI investigations. This information is requested annually as a requirement for data access.
Principal Investigator | |
Principal Investigator's Name: | SUHAIL AHMAD DAR |
Institution: | Aligarh Muslim University |
Department: | DEPARTMENT OF PSYCHOLOGY |
Country: | |
Proposed Analysis: | Lead Investigators: 1. Suhail Ahmad Dar, Research Scholar, Dept. Of Psychology, Aligarh Muslim University, India. 2.Umer Jan Ganai, Research Scholar, Dept. Of Humanities and Social Sciences, Indian Institute Of Technology, Kanpur, India. 3.Rameez Ahmad Dar, Dept. Of Computer Sciences and Engineering Tezpur University, Assam, India. Research Proposal The purpose is to do a retrospective longitudinal Study wherein people already diagnosed for Alzheimer’s will be evaluated for their Structural and Cognitive Functioning before their actual diagnosis. We will perform Particle Swarm Intelligence supported by Kernel Neural Networks and Support Vector Machine and simultaneously we will be performing Convolutional Neural Networks on the Time series Data that too on different Biomarkers consisting of Magnetic Brain Imaging, Positron Emission Tomography, Neuropsychological Data like MMSE, ADAS, CDR. This will be done at different time periods before the actual diagnosis i.e, 3-Years, 2-Year and 1-Year respectively. We will have two groups for MCI and these groups will be created on the Neuroimaging Database i.e, Voxel Based Morphometry and Volumetric Morphometry. The Study consists of 4 Stages-These are, • At Stage 1, Model Comparison (I.e, between PSO+KNN+SVM and CNN+LSTM) will be done on MCI (Volumetric Morphometric) group and MCI (Voxel Based Morphometric) group respectively. Each Model will undergo ROC, AUC analysis for Model Accuracy rate. The prediction models will include data inputs from MRI, PET, Clinical Assessment. This will be achieved through Time Series Data for 3-Years. • At stage 2, We will utilize the most predictive Model to compare the MCI group which involves Voxel based Morphometry with the MCI group which involves Volumetric Morphometry. The aim would be to see which kind of Morphometry performs better while predicting the Disease. • At Stage 3, the study will include the prediction of Alzheimer’s at 3-Years, 2-Years and 1-year separately. The Model with highest Accuracy at Stage 1, will be put into perspective here and We will be exploring how many Years before the actual diagnosis would the prediction be more successful. • At stage 4, the study will focus on the Classifiers with the highest predictive capability and that can be achieved through some statistical Ranking analysis i.e, Gini Index, T Score, Statistical Dependency, Fisher Criterion. Therefore, the current study proposes a multimodal retrospective time series design with the following contributions : to compare the predictive power of the volumetric and voxel based brain imaging scans for the prediction of Alzheimer’s disease; At Stage 1, multiple deep learning algorithms will be used and the model which stands out on accuracy in predicting the Alzheimer’s shall be used in the succeeding experiments; in experiment 3, the prediction rate will be examined retrospectively i.e, 3-Years, 2-Years and subsequently 1-Year before the Actual diagnosis, therefore the focus of this study is to evaluate the Year before the Actual diagnosis that displays the highest predictive rate for Alzheimer’s. That way we can increase the accuracy and avoid the unnecessary biases in the prediction. At stage 4, the classifiers will be separately analyzed for their contribution in diagnosing the Alzheimer’s and that can be achieved by running some statistical tests. |
Additional Investigators | |
Investigator's Name: | Umer Jon Ganai |
Proposed Analysis: | Lead Investigators: 1. Suhail Ahmad Dar, Research Scholar, Dept. Of Psychology, Aligarh Muslim University, India. 2.Umer Jan Ganai, Research Scholar, Dept. Of Humanities and Social Sciences, Indian Institute Of Technology, Kanpur, India. 3.Rameez Ahmad Dar, Dept. Of Computer Sciences and Engineering Tezpur University, Assam, India. Research Proposal The purpose is to do a retrospective longitudinal Study wherein people already diagnosed for Alzheimer’s will be evaluated for their Structural and Cognitive Functioning before their actual diagnosis. We will perform Particle Swarm Intelligence supported by Kernel Neural Networks and Support Vector Machine and simultaneously we will be performing Convolutional Neural Networks on the Time series Data that too on different Biomarkers consisting of Magnetic Brain Imaging, Positron Emission Tomography, Neuropsychological Data like MMSE, ADAS, CDR. This will be done at different time periods before the actual diagnosis i.e, 3-Years, 2-Year and 1-Year respectively. We will have two groups for MCI and these groups will be created on the Neuroimaging Database i.e, Voxel Based Morphometry and Volumetric Morphometry. The Study consists of 4 Stages-These are, • At Stage 1, Model Comparison (I.e, between PSO+KNN+SVM and CNN+LSTM) will be done on MCI (Volumetric Morphometric) group and MCI (Voxel Based Morphometric) group respectively. Each Model will undergo ROC, AUC analysis for Model Accuracy rate. The prediction models will include data inputs from MRI, PET, Clinical Assessment. This will be achieved through Time Series Data for 3-Years. • At stage 2, We will utilize the most predictive Model to compare the MCI group which involves Voxel based Morphometry with the MCI group which involves Volumetric Morphometry. The aim would be to see which kind of Morphometry performs better while predicting the Disease. • At Stage 3, the study will include the prediction of Alzheimer’s at 3-Years, 2-Years and 1-year separately. The Model with highest Accuracy at Stage 1, will be put into perspective here and We will be exploring how many Years before the actual diagnosis would the prediction be more successful. • At stage 4, the study will focus on the Classifiers with the highest predictive capability and that can be achieved through some statistical Ranking analysis i.e, Gini Index, T Score, Statistical Dependency, Fisher Criterion. Therefore, the current study proposes a multimodal retrospective time series design with the following contributions : to compare the predictive power of the volumetric and voxel based brain imaging scans for the prediction of Alzheimer’s disease; At Stage 1, multiple deep learning algorithms will be used and the model which stands out on accuracy in predicting the Alzheimer’s shall be used in the succeeding experiments; in experiment 3, the prediction rate will be examined retrospectively i.e, 3-Years, 2-Years and subsequently 1-Year before the Actual diagnosis, therefore the focus of this study is to evaluate the Year before the Actual diagnosis that displays the highest predictive rate for Alzheimer’s. That way we can increase the accuracy and avoid the unnecessary biases in the prediction. At stage 4, the classifiers will be separately analyzed for their contribution in diagnosing the Alzheimer’s and that can be achieved by running some statistical tests. |
Investigator's Name: | Rameez Ahmad Dar |
Proposed Analysis: | Lead Investigators: 1. Suhail Ahmad Dar, Research Scholar, Dept. Of Psychology, Aligarh Muslim University, India. 2.Umer Jan Ganai, Research Scholar, Dept. Of Humanities and Social Sciences, Indian Institute Of Technology, Kanpur, India. 3.Rameez Ahmad Dar, Dept. Of Computer Sciences and Engineering Tezpur University, Assam, India. Research Proposal The purpose is to do a retrospective longitudinal Study wherein people already diagnosed for Alzheimer’s will be evaluated for their Structural and Cognitive Functioning before their actual diagnosis. We will perform Particle Swarm Intelligence supported by Kernel Neural Networks and Support Vector Machine and simultaneously we will be performing Convolutional Neural Networks on the Time series Data that too on different Biomarkers consisting of Magnetic Brain Imaging, Positron Emission Tomography, Neuropsychological Data like MMSE, ADAS, CDR. This will be done at different time periods before the actual diagnosis i.e, 3-Years, 2-Year and 1-Year respectively. We will have two groups for MCI and these groups will be created on the Neuroimaging Database i.e, Voxel Based Morphometry and Volumetric Morphometry. The Study consists of 4 Stages-These are, • At Stage 1, Model Comparison (I.e, between PSO+KNN+SVM and CNN+LSTM) will be done on MCI (Volumetric Morphometric) group and MCI (Voxel Based Morphometric) group respectively. Each Model will undergo ROC, AUC analysis for Model Accuracy rate. The prediction models will include data inputs from MRI, PET, Clinical Assessment. This will be achieved through Time Series Data for 3-Years. • At stage 2, We will utilize the most predictive Model to compare the MCI group which involves Voxel based Morphometry with the MCI group which involves Volumetric Morphometry. The aim would be to see which kind of Morphometry performs better while predicting the Disease. • At Stage 3, the study will include the prediction of Alzheimer’s at 3-Years, 2-Years and 1-year separately. The Model with highest Accuracy at Stage 1, will be put into perspective here and We will be exploring how many Years before the actual diagnosis would the prediction be more successful. • At stage 4, the study will focus on the Classifiers with the highest predictive capability and that can be achieved through some statistical Ranking analysis i.e, Gini Index, T Score, Statistical Dependency, Fisher Criterion. Therefore, the current study proposes a multimodal retrospective time series design with the following contributions : to compare the predictive power of the volumetric and voxel based brain imaging scans for the prediction of Alzheimer’s disease; At Stage 1, multiple deep learning algorithms will be used and the model which stands out on accuracy in predicting the Alzheimer’s shall be used in the succeeding experiments; in experiment 3, the prediction rate will be examined retrospectively i.e, 3-Years, 2-Years and subsequently 1-Year before the Actual diagnosis, therefore the focus of this study is to evaluate the Year before the Actual diagnosis that displays the highest predictive rate for Alzheimer’s. That way we can increase the accuracy and avoid the unnecessary biases in the prediction. At stage 4, the classifiers will be separately analyzed for their contribution in diagnosing the Alzheimer’s and that can be achieved by running some statistical tests. |
Investigator's Name: | Rameez Ahmad Dar |
Proposed Analysis: | Lead Investigators: 1. Suhail Ahmad Dar, Research Scholar, Dept. Of Psychology, Aligarh Muslim University, India. 2.Umer Jan Ganai, Research Scholar, Dept. Of Humanities and Social Sciences, Indian Institute Of Technology, Kanpur, India. 3.Rameez Ahmad Dar, Dept. Of Computer Sciences and Engineering Tezpur University, Assam, India. Research Proposal The purpose is to do a retrospective longitudinal Study wherein people already diagnosed for Alzheimer’s will be evaluated for their Structural and Cognitive Functioning before their actual diagnosis. We will perform Particle Swarm Intelligence supported by Kernel Neural Networks and Support Vector Machine and simultaneously we will be performing Convolutional Neural Networks on the Time series Data that too on different Biomarkers consisting of Magnetic Brain Imaging, Positron Emission Tomography, Neuropsychological Data like MMSE, ADAS, CDR. This will be done at different time periods before the actual diagnosis i.e, 3-Years, 2-Year and 1-Year respectively. We will have two groups for MCI and these groups will be created on the Neuroimaging Database i.e, Voxel Based Morphometry and Volumetric Morphometry. The Study consists of 4 Stages-These are, • At Stage 1, Model Comparison (I.e, between PSO+KNN+SVM and CNN+LSTM) will be done on MCI (Volumetric Morphometric) group and MCI (Voxel Based Morphometric) group respectively. Each Model will undergo ROC, AUC analysis for Model Accuracy rate. The prediction models will include data inputs from MRI, PET, Clinical Assessment. This will be achieved through Time Series Data for 3-Years. • At stage 2, We will utilize the most predictive Model to compare the MCI group which involves Voxel based Morphometry with the MCI group which involves Volumetric Morphometry. The aim would be to see which kind of Morphometry performs better while predicting the Disease. • At Stage 3, the study will include the prediction of Alzheimer’s at 3-Years, 2-Years and 1-year separately. The Model with highest Accuracy at Stage 1, will be put into perspective here and We will be exploring how many Years before the actual diagnosis would the prediction be more successful. • At stage 4, the study will focus on the Classifiers with the highest predictive capability and that can be achieved through some statistical Ranking analysis i.e, Gini Index, T Score, Statistical Dependency, Fisher Criterion. Therefore, the current study proposes a multimodal retrospective time series design with the following contributions : to compare the predictive power of the volumetric and voxel based brain imaging scans for the prediction of Alzheimer’s disease; At Stage 1, multiple deep learning algorithms will be used and the model which stands out on accuracy in predicting the Alzheimer’s shall be used in the succeeding experiments; in experiment 3, the prediction rate will be examined retrospectively i.e, 3-Years, 2-Years and subsequently 1-Year before the Actual diagnosis, therefore the focus of this study is to evaluate the Year before the Actual diagnosis that displays the highest predictive rate for Alzheimer’s. That way we can increase the accuracy and avoid the unnecessary biases in the prediction. At stage 4, the classifiers will be separately analyzed for their contribution in diagnosing the Alzheimer’s and that can be achieved by running some statistical tests. |