Mapping Fcuntional Brain Connectivity Alterations in Alzhiemier's Disease with Machine Learning
Title: Mapping Fcuntional Brain Connectivity Alterations in Alzhiemier's Disease with Machine Learning
DNr: Berzelius-2023-41
Project Type: LiU Berzelius
Principal Investigator: Jiawei Sun <jiawei.sun@ki.se>
Affiliation: Karolinska Institutet
Duration: 2023-03-03 – 2023-10-01
Classification: 30105
Keywords:

Abstract

The aims of this project are: 1. Detect which functional brain connections show nonlinear changes over the course of aging and in different stages of AD. 2. Establish how these changes are associated with amyloid and tau pathology 3. Determine whether mechanisms based on synaptic pruning explain the nonlinear changes in brain connectivity in AD. 4. Apply the methods developed in this project in clinical practice to predict cognitive decline and dementia. Our intention is to undertake the development of a deep learning method based on Graph Attention Networks. To achieve this, we plan to utilize various deep learning toolboxes including TensorFlow, PyTorch, and others in our projects. As this task involves computationally intensive operations, we require a robust server to support our calculations. Therefore, we request a suitable server to facilitate our work.