QuickSeg aims to develop a deep learning-based brain segmentation system capable of accurately processing low to middling quality PET scans. By quantifying average pixel brightness in various brain regions related to glucose measurements, the system will aid medical professionals, particularly neurologists and radiologists, in detecting neurological disorders and analyzing brain metabolism. This implementation will have broader implications for neuroscience research, advancing our understanding of brain function and connectivity.


Community Benefit

Removing portions of a brain scan from the image is useful for any study using computers for analysis. By removing the skull and background noise, QuickSeg will improve the accuracy and efficiency of machine learning algorithms attempting to classify disease in a given brain. Additionally, QuickSeg measures relative glucose levels in brains, allowing researchers to study the relationship between glucose and disease.


Team Members

Mohannad Darwish

Kevin Infante 

Kyle Palmer

Justin Rivera

Dustin Sherwood



Dr. Rui Tao