Medical Imaging

Diagnostic Imaging research at UBC Okanagan, BC Cancer Agency – Centre for the Southern Interior, and Interior Health is clinically oriented with focus on quality control, patient dose optimization, and radiation safety.

Quality Control

Characterizations of image quality for CT with iterative reconstruction algorithm (IRA)

Iterative reconstruction algorithms (IRA) are now widely available on many modern CT systems. We are investigating the feasibility of routinely and objectively characterizing the image quality of CT with IRA. Conceptually, noise-equivalent-quanta (NEQ) potentially can be a good metric of image quality assessment for the whole system. This study investigates and develops the methodology for NEQ assessment of IRA capable CT systems using American College of Radiology (ACR) accreditation phantom

Quality Control Charts and Graphs

CT radiation profile width measurement using CR imaging plate raw data

Measuring the Computed Tomography (CT) collimation width is a straightforward quality control (QC) measure that impacts both image quality and patient dose. Many jurisdictions and accreditation agencies require the profile to be assessed at acceptance and as part of routine quality control. Many sites use disposable film or computed radiography (CR) plates along with careful CR reader modifications. We have shown that using raw CR data allows CT collimation measurements to be performed without requiring CR reader adjustments. Most CT facilities have access to computed radiography (CR) systems and performing CT collimation profile assessments using CR requires no additional equipment or consumables, unlike film alternatives.

CT radiation data chart

People

Thor Bjarnason, Joseph Yang, Rebecca Feldman

Population Dose Optimization – Diagnostic Reference Levels

Diagnostic reference levels (DRLs) are used to optimize the trade-off between patient dose and image quality. A DRL for a given examination is established by performing a large survey of patient dose indicators for a given patient weight (or a narrow range of weights). By looking at the histogram of the results, outliers can be identified for follow-up. The upper DRL is typically defined as the 75th percentile of the distribution of dose indices. If a site uses doses consistently above the upper DRL it means it is worth following-up with the site to see if they can achieve diagnostic quality images at a lower dose. A lower DRL can be set at the 10th percentile to identify sites that might be using too little dose and could benefit from better image quality comparable to their peers. For large health regions with limited resources, the DRL approach is an efficient way to optimize patient dose.

CT DRLs

CT Chart DataComputed tomography is the largest contributor to manmade population dose. Reducing individual patient dose in aggregate reduces population dose and the corresponding potential stochastic effects. Computed tomography quality improvement audits are performed as part of the medical physics semi-annual quality control process. In an examination of CT head doses, the collective dose to patients was 70 person-Sv/yr prior to protocol review and an overall reduction to the median dose length product of 19 % was achieved, resulting with a collective dose reduction of 13 person-Sv/yr for patient population receiving head CTs. The picture below shows the box plot representation of doses for a specific (anonymized) site before (black) and after (grey) the audit. Ongoing work is being performed for for Chest/Abdomen/Pelvis, L-spine, and Abdomen/Pelvis exams.

Novel Applications

Computed Radiography Plates as dosimeters

CR plates are commonly used for digital x-ray detection, and are convenient alternatives to conventional film technology. Before a CR plate can be used as both a quantitative and qualitative dosimeter, signal fading of the underlying photostimulable phosphor technology needs to be characterized using a model of a sum of exponential decays that allows the determination of decay constants from known exposures and measurement delays over a long period of time. This complex signal fading characterisation is needed in order to determine the cumulative absorbed dose to the CR plate accurately over the measurement timeframe. A paperclip was taped to the front of the CR plate which was then affixed to the wall in a diagnostic CT control booth without a shielding door leading to the main CT room for 2 months. Two shadows of the paper clip were captured in the image, giving indication of the directions of the two primary sources of scatter into the control booth. Quantitative analysis of the CR plate leads to the dose measurements in agreement with ionization chamber measurements. Using a CR plate as an area dosimeter has the advantage of providing a visual tool in understanding the presence and amount of radiation scatter.

Signal Decay and Model Fit