Dynamic Image for 3D MRI Image Alzheimer's Disease Classification
This work provides a more efficient method for Alzheimer's disease classification from MRI scans, which could benefit clinical diagnosis and research by reducing computational burden.
This paper addresses the computational expense of 3D CNNs for Alzheimer's disease classification by transforming 3D MRI volumes into 2D images using approximate rank pooling. This approach resulted in a 9.5% improvement in classification accuracy and reduced training time by 80% compared to baseline 3D models.
We propose to apply a 2D CNN architecture to 3D MRI image Alzheimer's disease classification. Training a 3D convolutional neural network (CNN) is time-consuming and computationally expensive. We make use of approximate rank pooling to transform the 3D MRI image volume into a 2D image to use as input to a 2D CNN. We show our proposed CNN model achieves $9.5\%$ better Alzheimer's disease classification accuracy than the baseline 3D models. We also show that our method allows for efficient training, requiring only 20% of the training time compared to 3D CNN models. The code is available online: https://github.com/UkyVision/alzheimer-project.