Nuclear Medicine Physics Laboratory
About us
The Nuclear Medicine Physics Laboratory is managed by Ran Klein, PhD. It conducts innovative research and development and also provides clinical support to the Department of Nuclear Medicine at The Ottawa Hospital. Students in the lab are affiliated with Carleton University, Departments of Physics and Department of Systems and Computer Engineering. We are also closely affiliated with the University of Ottawa, Nuclear Cardiology program. Our research is primarily in the domains of:
- Medical image analysis
- Quantification of physiologic function
- Computer aided diagnosis
- Lesion detection
- Organ segmentation
- Myocardial blood flow quantification
- Motion detection and correction
- Quantitative single photon emission computed tomography (SPECT)
- Positron emission tomography (PET)
Highlighted Projects
Disease Modelling PET Toolbox
This toolbox has 3 main products. (1) DiscoveryDR an over the network data retrieval tool to extract raw PET and CT data from a Discovery 600/690/710 PET/CT System. (2) DiscoveryIR applies clinically available GE PET image reconstruction algorithms on real clinical data. (3) vDiscovery is an analytical time-of-flight PET simulation package used to generate projection planes of simulated objects considering attenuation and other scanner effects. Student: Hanif Gabrani-Juma, B.Eng, MASc
Artifical PET/CT Observer
Using analytical methods and machine-learning we are developing artifical intelligence (AI) for computer aided diagnosis (CAD) of FDG PET-CT studies. In collaboration with Hermes Medical Solutions we are enabling collaboraion between human and machine observers using Hybrid3DTM to improve diagnostic accuracy, automatically generate templated clinical reports, and to provide active learning feedback to our AI. Student: O Salman, PhD Candidate
Motion Correction in raw PET
Patient and organ motion during PET data acquisition are a key element of image degredation that can adversely impact the clinical utility of the exam. Motion tracking during acquistion, can be incorporated to the image reconstruction process to cancel out motion effects. We are using dedicated hardware and data-driven methods to measure patient and repiratory motion in the context of dynamic cardiac PET for myocardial blood flow quantification in collaboration with the University of Ottawa Heart Institute, National Cardiac PET Centre. Student: S Manwell PhD Candidate (co-supervised by Dr Robert deKemp and Tong Xu)
DICOM Converter for MatLab
Working with DICOM series can be slow and awkward, to speed things up, we frequently convert DICOM series to a matlab file that contains the entire binary image data, dicom header and a summary header. This has the added benefit of significantly reducing disk space usage.