8th Baltic Congress of Radiology

Radioprotection and management

oral abstract

Medical imaging represents the largest source of radiation exposure from artificial sources. Since high level of ionizing radiation has an adverse effect on human health, radiation dose form medical exposure should be kept ‘‘as low as reasonably achievable’’ (ALARA). Digital radiography has gained popularity and replaced film-screen radiography in most developed countries since it offers many advantages over the conventional systems such as wide dynamic range and the post-processing capabilities. However, the images and reconstructions of digital radiography systems vary depending on the manufacturer and model. Thus, optimization is necessary for each digital radiography equipment unit. Therefore, the application of the optimization principle in digital radiography is key to balance acceptable image quality with patient dose. The aim of this study was to assess the impact of varying acquisition parameters on image noise and contrast with the objective of establishing optimal pelvic imaging protocol for different digital X-Ray units.

: Eight digital radiography equipment from four different manufacturers (General Electronics, Siemens, Shimadzu and Philips) operating at Vilnius University Hospital Santaros Klinikos, Vilnius, Lithuania were included in this study. A male pelvis anthropomorphic phantom (CIRS MODEL 801-P, USA) was used to simulate a standard size patient under AP pelvic diagnostic X-ray imaging. Images were acquired using fixed exposure (~80 kV, 10 mAs) in order to compare the performance of all radiology systems. For each protocol, a set of images were obtained in automatic exposure control (AEC) mode using the standard pelvic protocol set by each manufacturer as well as different tube voltages from 65 to 90 kV at 5 kV step interval. Moreover, the effect of filtration on image quality and patient dose was assessed by repeating the previously performed exposures with added 0.1 mm Cu. Dose following each exposure was reported as Dose area product (DAP) collected from each system and as Entrance surface dose (ESD) which was measured by multiplying incident air kerma, measured with a solid state sensor (Piranha RTI Dose Probe) placed on the surface of the phantom, by a backscattering factor corresponding to each beam quality. Image quality assessment was carried out objectively by measuring contrast-to-noise ratio which is related to the contrast or signal difference between an object of interest and the image background. CNR was measured using the open source software ImageJ while taking the sacrum and soft tissue as object of interest and background respectively. Optimization strategy used in this study was based on analysing Figure of merit (FOM), defined as contrast-to-noise ratio (CNR) squared per unit dose. FOM was used to evaluate the efficiency of the X-ray system in producing a high CNR image with a low ESD.

A total of 120 PA pelvis X-Ray images obtained from 8 different manufacturers were included in this study. The standard AEC pelvic protocol (kV, mA, FOV, AEC) used in the clinical setting differs for each manufacturer. Tube voltage were set at 75 kV for Shimadzu, 77 kV for Siemens, 80 kV for GE and 85 kV for Philips. Direct comparison of these three units at a fixed radiation exposure (~80 kV; 10 mAs) showed a wide difference in CNR (ranging between 0.4 for Shimadzu RadSpeed and 2 for GE Definium) and exposure dose (DAP ranging from 5.3 and 9 dGy*cm2) across all units. Even different equipment from the same manufacturer had a significant difference in image quality and registered doses at the same exposure parameters. Additionally, the results indicated that DAP measurements set by all manufacturer can be provide an accurate estimation of patient exposure since ESD and DAP values agreed well in all cases. For pelvic protocol in AEC, both image quality (CNR) and radiation dose (ESD and DAP) decreased with increasing tube voltage for all x-ray units. However, percentage change in ESD and DAP was different across each unit. The results showed that filtration with 0.1 mm Cu was a favorable method of dose optimization while preserving image quality. By adding filtration to the standard protocol, image quality only decreased by 2-7.2% while patient dose decreased by 12-40.4%. Dose reduction while maintaining image diagnostic quality, was achieved by using minimized dose and maximized FoM. This optimization strategy indicating that clinical protocol set by manufacturers for pelvic x-ray examinations could and should be adjusted.

An anthropomorphic phantom was successfully used in patient dose optimization of digital radiographic examinations of the pelvis. The results indicated that optimization should be individualized for each equipment due to significant difference in their performances. Addition of 0.1 mm Cu filtration resulted in dose reduction with comparable image quality for all equipment. Evaluation of FOM showed that optimal dose-image quality necessitates a unique change in kV for each x-ray equipment while for some systems simple addition of filtration is enough. In this study, the optimization strategy involved the assessment of physical parameters (CNR, ESD, DAP). Finally, even though visual evaluation of images by radiologist was not included, it should be done before implementation of new protocols to the clinical practice.

A male pelvis anthropomorphic phantom (CIRS, USA) was used to simulate a standard patient under AP pelvic diagnostic X-ray imaging. Eight digital radiography equipment from four different manufacturers were compared in this study. Image were acquired using the three different protocol using fixed exposure and AEC mode with different tube voltages. An anthropomorphic phantom was successfully used in patient dose optimization of digital radiographic examinations of the pelvis.

1242