Image source: link
Project completed as part of Udacity AI for Healthcare course (2020).
U-net model is applied to MRI brain scan to segment volume of the hippocampus. Measuring hippocampal volume is useful in monitoring progression of Alzheimer's disease, however it is time consuming and challenging to do by hand. Automated AI segmentation offers fast and accurate 3D volume results for interpretation by radialogist.
Sample Result DICOM with Hippocampal Volume Segmentation:
Section 1: Exploratory Data Analysis
Section 2: Model training
Section 3: Execute segmentation model as part of mock clinical network and publish findings to medical viewer.