Central Nervous System

Brain Hemorrhage Project – Real-time Intracerebral Hemorrhages (ICH) Detection, Localization, and Quantification

We are developing a deep learning tool to automatically detect intracranial hemorrhage on Head CT while providing both 3D hemorrhage volume and expansion prediction estimate. Should ICH be detected following an emergent scan in the Emergency department, the software would then automatically alert the radiologist for immediate review. We hope this tool will be able to help physicians and healthcare providers gain a more accurate assessment of a patient’s hemorrhage while also improving turn-around time to improve patient outcomes.

Brain Extraction and Segmentation Project – Comparison to the Gold Standard

Brain extraction, the removal of the brain from non-brain tissue on imaging, is typically the first imaging preprocessing step of neuroimaging analysis. Current tools widely used in the field are either fast but less accurate than manual methods, or accurate but computationally taxing. We have developed an automated deep learning method capable of fast and highly accurate brain extraction and segmentation of tissue types on imaging.

Brain Tumor Projects – The Cancer Genome Atlas (TCGA) Dataset

The Cancer Genome Atlas is a genomics program is a research initiative that generates a vast amount of genomic and clinical data. Using Glioblastoma (GBM) and Low Grade Glioma (LGG) datasets from this program, we are creating a tool that will ultimately predict the survival rate of the patient in years by assessing MRIs. We hope this tool will increase the ability for a clinician to diagnose, treat, and prevent cancer in a more accurate and efficient manner.