This is a project which is currently making use of HPC facilities at Newcastle University. It is active.
For further information about this project, please contact:
High-resolution histological images from 200 DLB cases and age-matched controls will be analysed using a custom convolutional neural network with human-level accuracy to detect neurons. The model will be optimised for the heterogeneity of human tissue and to automate detection of neuronal density and alpha-synuclein aggregates. Large-scale quantitative datasets will be generated across multiple temporal lobe regions and analysed using appropriate statistical models. The student will receive training in the preparation, imaging, and neuropathological assessment of human post-mortem brain tissue. Parallel AI-based analysis of NHP viral-tracing material will be used to map memory-network connectivity, integrating both datasets.
We will primarily use python-based software to train convolutional neural networks such as YOLO and SAM. We will also use comet to deploy trained models to make detection and classification predictions on large histological datasets.