Deep learning-based multitask and multimodal study to assess the progressive nature of Alzheimer's disease

Alzheimer’s disease (AD) is a progressive neurodegenerative disease that is irreversible. Theprogressive nature of the disease affects the ability to carry out everyday activities in an early stage and thelater severe stages; the person is incapable of independent living. Early detection of AD will help to reducethe cognitive impairment and thus slow down the progressive nature of the disease by providing socialengagement and stimulating activities both mentally and physically [1]. The brain tissues startdegeneration much earlier than the clinical symptoms appear. So it is crucial to look at brain imaging toidentify the early-onset of AD. So, this project aims to develop a customized optimal diagnostic modelfor classifying the AD stages inclusive of early on-set from structural Magnetic Resonance Imaging. Development of adaptive pre-processing algorithms for structural Magnetic Resonance Imaging (MRI) To propose an effective segmentation method using attention maps to quantify the volumetricmeasurement of AD during all stages from MRI data. To extract optimal discriminative bio-imaging markers using the deep learning methodology. To develop multi-task and multimodal 3D deep neural network models for the classification of AD. To develop an ensemble of regression models to predict the prognostic index of AD