Timothy Peter Szczykutowicz PhD, Presenter: Equipment support, General Electric Company Research Grant, Siemens AG

Myron Andrew Pozniak MD, Abstract Co-Author: Stockholder, Cellectar Biosciences, Inc

Frank N. Ranallo PhD, Abstract Co-Author: Grant, General Electric Company

Patient positioning in CT is critical for obtaining the lowest possible imaging dose and minimizing artifact level. Traditionally, CT technologists are instructed to align the geometric center of patients with the scanner’s isocenter. We propose an alternative positioning and evaluate three metrics for position determination.

Localizer images were taken from 184 clinical abdominal patients and one anthropomorphic phantom at our institution and three centering determination metrics were applied to each. The optimality of the metrics was based on which correlated best with positioning resulting in good noise uniformity and low beam hardening artifact level determined using the anthropomorphic phantom. The first metric uses the “center of mass” of the profile from the localizer image. For each detector row, the profile is extracted, thresholded, and the center of mass is calculated. The second method thresholds each line of the localizer image and the edges of the remaining profile are extracted and used to determine the geometric center of the patient. The third method calculates the center of mass of the highest 3% of the projection values within the localizer.

On average, the geometric center metric provided the most “anterior” offcentering measurement, followed by the “center of mass” (0.2 cm lower relative to the geometric center) and then the COM of the highest 3% of the attenuation profile (1.7 cm lower relative to the geometric center). The third approach may be the most clinically relevant approach because it is directly correlated with how well the most attenuating part of the patient is positioned relative to iso-center. Having the most attenuating part of the patient aligned with isocenter in most cases ensures the lowest level of image noise non-uniformity and beam hardening.

Due to patient tissue inhomogeneity, patient positioning determination metrics based on the geometric center of the patient will not accurately provide information on optimal patient positioning. Tissue inhomogeneity within patients must be taken into account in order for position metrics to act as surrogates for optimal patient positioning.

Dose monitoring companies are offering tools to aide in determining proper patient positioning, however, these tools may not be surrogates for optimal positioning due to patient tissue inhomogeneity.

Szczykutowicz, T, Pozniak, M, Ranallo, F, A Comparison of Three Methods for Measuring Patient Positioning from Localizer Imaging in CT: Which Correlates Best with Optimal Image Quality?.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14004979.html