Using structural MRI images, a multi-stream convolutional neural network can classify progressive MCI in Alzheimer's disease.

PROJECT TITLE :

A Multi-Stream Convolutional Neural Network for Classification of Progressive MCI in Alzheimer’s Disease Using Structural MRI Images

ABSTRACT:

It is essential to perform an early diagnosis of Alzheimer's disease and its prodromal stage, also referred to as mild cognitive impairment (MCI). This is due to the fact that some patients who have progressive MCI will eventually develop Alzheimer's disease. In order to differentiate between stable MCI and progressive MCI, we suggest using a multi-stream deep convolutional neural network that is fed with patch-based imaging data. In the first step of this process, we use a multivariate statistical test to compare MRI images of Alzheimer's disease patients with those of cognitively normal subjects in order to locate distinct anatomical landmarks. Following the identification of these landmarks, patches are extracted and then fed into the proposed multi-stream convolutional neural network for the purpose of MRI image classification. Next, we train the architecture in a separate scenario using samples from Alzheimer's disease images, which are anatomically similar to the ones of progressive MCI, and cognitively normal images to compensate for the lack of progressive MCI training data. These samples come from both Alzheimer's disease and cognitively normal images. In the final step of this process, we transfer the trained model weights to the proposed architecture in order to fine-tune the model by using progressive MCI and stable MCI data. With an F1-score of 85.96%, the results of our experiments on the ADNI-1 dataset indicate that our method performs better than other methods currently used for the classification of MCI.

Did you like this research project?

To get this research project Guidelines, Training and Code... Click Here