OMPI Seminar: Mehan Haidari and Rolf Clackdoyle

Printer-friendly version
Date: 
Thursday, January 21, 2016

Time: 3:30-5:00 pm - Thursday 21 January 2016. Refreshments start at 3:15 pm.

Location: Room RPB 205 (boardroom), Health Canada, 775 Brookfield Road.

Program:

1. Retrospective dosimetric Monte Carlo study for permanent implant prostate brachytherapy at the Ottawa Hospital Cancer Centre

By Mehan Haidari, Carleton University.

Abstract: Current clinical dose calculation methods for low dose-rate (LDR) prostate brachytherapy are performed following the protocols defined by the AAPM Task Group report no. 43 (TG43) formalism. Under the TG-43 formalism, absorbed dose is calculated in a homogeneous water phantom, ignoring the effects of tissue heterogeneities. Additionally, dose contributions from each radionuclide are considered independently, ignoring interseed attenuation and scattering effects. Using Monte Carlo (MC) techniques, these limitations inherent to TG-43 are overcome. This talk will investigate the MC technique used to calculate dose distributions for a cohort of patients that received LDR prostate brachytherapy at the Ottawa Hospital Cancer Center. Furthermore, some work done in linking patient outcomes with these more accurate dose calculation models will be discussed.


2. Data consistency conditions and their relevance to medical imaging

By Rolf Clackdoyle, Hubert Curien Laboratory in France, and Physics Deartment at Carleton University

Abstract: X-ray Computed Tomography (CT), Positron Emission Tomography (PET), and Single Photon Emission Computed Tomography (SPECT) scanners gather "projections" at different orientations around the patient, and these projections are assembled by the reconstruction algorithm to form a three-dimensional image of the corresponding physical parameter (electron density, positron-emitting isotope, gamma-emitting isotope respectively).

These projections gather largely independent information, which is why so many of them are measured. However, there is a small amount of redundancy (or overlap of information) between projections. This redundancy can be described precisely using mathematical equations referred to as data consistency conditions (DCC).  The DCC can be used to remove undesired systematic effects (such as patient motion, to name one of many examples) from the measurements before image reconstruction. Very recently there has been progress in identifying DCC for fanbeam and conebeam imaging geometries which are particularly suitable for CT.  One such example of new fanbeam DCC will be presented with a toy problem to illustrate the approach and efficacy in removing systematic effects, even in the presence of (zero-mean) noisy data.