Precise and rapid whole-head segmentation from magnetic resonance images of older adults using deep learning
Developed a deep learning model that specializes in the segmentation of brain images of older adults. Many current models are trained on younger adults and may not apply to an older population.
Principal Investigator: Dr. Ruogu Fang
Department: Biomedical Engineering
Years with Project: 1 year
Responsibilities
-
Wrote a program to visualize 2D slices of segmentations, with the ability to color different tissue types different colors.
-
Wrote a program to calculate the accuracy of segmentations produced by the model.
Broader Impact
Brain segmentations can help with a variety of treatments, including non-invasive brain stimulation. These treatments are often used in older adults for age-related disorders. By creating a model specific to the older population, these segmentations and treatments can be more accurate.