Paul Sharoubeem

Medical Data Scientist @University of Pittsburgh School of Medicine

Pittsburgh, PA, US

Designed top of industry Neuroimaging based computational analysis pipelines to prepare Magnetic Resonance and PET Scans for Research and downstream physician analysis. Built custom Python Scripts utilizing top neuroimaging soft packages such as Antspy, Nibabel, Sklearn, Freesurfer, and Pandas to preprocess DICOM files of PET and MR Scans into NIFTI format, Co-Registering PET and Structural MR neuroimage scans, Parcelization of MR Structures, automating motion correction and decay correction, and computing GTM partial volume correction to build hospital wide statistical databases. Automated and enhanced antiquated pipelines using Docker and XNAT to build robust containers that resulted in an increase in analysis speed and lowering margin of errors that used to arose from manual neuroimaging analysis and user error, standardizing our department PET White Matter Hypersensitivity analysis.