Arjun Maniyedath MS, Presenter: Employee, Plexar Associates, Inc

Frank Dong PhD, Abstract Co-Author: Equipment support, Siemens AG Software support, Siemens AG

Kassem Soufan, Abstract Co-Author: Nothing to Disclose

David Rohler PhD, Abstract Co-Author: Employee, Plexar Associates, Inc

Sandra Simon Halliburton PhD, Abstract Co-Author: Research Grant, Koninklijke Philips NV Research Grant, Bayer AG

CT Protocol recommendation tools used to determine tube potential and tube current based on the specific clinical indication and patient size were developed independently for two scanner types using a time-consuming, manual process. The purpose of this work was to determine the utility of a comprehensive quantitative image quality metric based on Channelized Hotelling Model Observer (CHO) to more efficiently predict patient size-dose curves for a new CT scanner platform based on existing curves for a different scanner. 

Previously validated CT scanner-specific protocol recommendation tools are used in daily clinical practice to determine optimal tube potential and tube current settings based on patient size for the evaluation of thoracic aortic disease with CT. Patient size (water equivalent diameter [Dw]) and dose (CTDIvol) were determined for patients from each of two scanners (256-slice Philips Brilliance iCT [Scanner 1] and 64-slice Philips Brilliance 64 [Scanner 2]). Both scanners were evaluated for image quality using a phantom containing three sections (Dw = 12, 25 and 32 cm), at five dose levels. Each section contains a region with three 5 mm and 2 mm low contrast rods, a uniform region, and a region to measure rod contrasts in the image (Plexar Imaging phantom). This characterization incorporates a CHO based image quality metric and generates an Image Index curve for the scanners across a wide range of body sizes and dose levels. Using the size-dose curve determined from patient scanning on Scanner 1 and the Image Index curve determined from phantom scanning on both scanners, the required size-dose curve for Scanner 2 was predicted.

93 patients (Dw range: 20-37 cm) were included in the study. The predicted size-dose curve for Scanner 2 closely fit the actual curve across the patient size range (Figure). 

Scanner characterization using the ConvergeCT phantom and the Image Index metric enables the automated development of scanner and indication-specific protocol recommendation tools for new scanner platforms based on established practices.  

Scanner characterization using a dedicated phantom and quality metric provides an objective, automated process for adjusting proven protocols for new CT scanner platforms.

Maniyedath, A, Dong, F, Soufan, K, Rohler, D, Halliburton, S, Efficient Mapping of Protocol Selection Targets across Different CT Scanners Using Channelized Hotelling Observer Based Image Quality Metric.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14019407.html