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: | SHAIK BASHEERA |
Institution: | ACHARYA NAGARJUNA UNIVERSITY |
Department: | ELECTRONICS AND COMMUNICATION ENGINEERIG |
Country: | |
Proposed Analysis: | Alzheimer’s is Progressive dementia, early detection of AD is a challenging task, different computer-aided diagnosis techniques are adapted to diagnosis AD using neuroimaging prominently MRI Images. We develop a CNN model to assist Physician for early diagnosis of AD. Atrophy changes in the subcortical region of Brain MRI at the early stage of AD and normal subject are the same. Measuring the changes in gray matter is one of the primary features for diagnosis of AD, in our work we used segmented gray matter of the T2w MRI for early diagnosis of AD. In our experiment, we used 1820 T2w MRIs volumes those are collected from ADNI data set and generate 18017 numbers of slices. We further divided the dataset into 80% training and validation, 20% of data for testing. To test the system performance, we used different statistical parameters such as Accuracy, Precession, Recall, F1 Score. In this work, we used Binary and multi-class classification, during subject level classification we achieve 96.78% of accuracy in case of AD-CN-MCI, 93.58% of accuracy in case of AD-MCI, 100% of accuracy in case of AD-CN, 96.92% of accuracy in case of CN-MCI. The model is further trained with 10-fold cross-validation and got an accuracy of 92.92±3% in case of CN-MCI, 98±2% accuracy in case of CN-AD, 90±4% in case of AD-MCI, and 94.9±2% in case of CN-MCI-AD. Our proposed approach significantly improves the diagnosis of AD from MCI and CN while compared with the previous works and used for early diagnosis of AD. |
Additional Investigators | |
Investigator's Name: | M SATYA SAI RAM |
Proposed Analysis: | Alzheimer’s is Progressive dementia, early detection of AD is a challenging task, different computer-aided diagnosis techniques are adapted to diagnosis AD using neuroimaging prominently MRI Images. We develop a CNN model to assist Physician for early diagnosis of AD. Atrophy changes in the subcortical region of Brain MRI at the early stage of AD and normal subject are the same. Measuring the changes in gray matter is one of the primary features for diagnosis of AD, in our work we used segmented gray matter of the T2w MRI for early diagnosis of AD. In our experiment, we used 1820 T2w MRIs volumes those are collected from ADNI data set and generate 18017 numbers of slices. We further divided the dataset into 80% training and validation, 20% of data for testing. To test the system performance, we used different statistical parameters such as Accuracy, Precession, Recall, F1 Score. In this work, we used Binary and multi-class classification, during subject level classification we achieve 96.78% of accuracy in case of AD-CN-MCI, 93.58% of accuracy in case of AD-MCI, 100% of accuracy in case of AD-CN, 96.92% of accuracy in case of CN-MCI. The model is further trained with 10-fold cross-validation and got an accuracy of 92.92±3% in case of CN-MCI, 98±2% accuracy in case of CN-AD, 90±4% in case of AD-MCI, and 94.9±2% in case of CN-MCI-AD. Our proposed approach significantly improves the diagnosis of AD from MCI and CN while compared with the previous works and used for early diagnosis of AD. |