Usman Mahmood MS, Presenter: Nothing to Disclose

Yusuf Emre Erdi DSc, Abstract Co-Author: Nothing to Disclose

To determine low contrast detectability (LCD) in the pediatric patient while maintaining the noise magnitude and texture.

A CIRS liver phantom with three rows of 7 spherical targets, ranging from 10 mm to 2.5 mm, that are 5, 10, and 20 HU below the liver equivalent background was used to asses low contrast detectability. A Gammex 464 CT accreditation phantom was used to calculate CNR. The minimum observable spherical target was detected under appropriate viewing conditions. CNR was calculated according to the method described in the ACR CT phantom scanning instructions. The phantoms were scanned with the current pediatric CT protocol (80 mA, 10% iterative reconstruction). In order to maintain an equivalent level of noise in the reduced dose protocol (45 mA, 40% iterative reconstruction), a decrease in mA was accompanied by an increase in iterative reconstruction. Multiple levels of mA and iterative reconstruction between the current protocol and the reduced protocol have also been investigated.  

A minimum spherical target diameter of 6.3 mm was detectable on reconstructed images acquired with the current CT protocol. With the reduced dose protocols, a minimum spherical target diameter of 9.5 mm was detectable on the reconstructed images. CNR was 0.593 ± 0.006 for all protocols. For an equivalent noise magnitude and texture as the current CT protocol, CTDI was found to decrease by up to 43%.  

Optimization of CT protocols, while maintaining an equivalent noise magnitude and texture of CT images, leads to a loss of LCD. Furthermore, trying to optimize protocols based on the CT accreditation phantom alone may not be enough for optimum LCD.

Dose reduction achieved with a decreased mA and increased application of iterative reconstruction may result in a loss of LCD.

Mahmood, U, Erdi, Y, Impact of Iterative Reconstruction and Low Dose on Low Contrast Detectability in Pediatric Patients in CT.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14002849.html