Challenges for computed tomography of overweight patients

In morbidly obese patients, computed tomography frequently represents the only viable option for non-invasive imaging diagnostics. The aim of this study was to analyze the weight limits, dose and image quality with standard CT scanners and to determine the diagnostic value and dose with a dual source XXL mode. A total of 15 patients (average body weight 189.6±42 kg) were retrospectively identified who had been examined with the XXL mode. Of these patients 7 (average body weight 176.4±56 kg) had been examined using both the XXL and standard protocols allowing for an intraindividual comparison in this subcollective. Additionally 14 patients weighing between 90 and 150 kg (average 106.1±19 kg) examined with standard protocols were included as references. Dose, image noise and subjectively assessed image quality (rating scale 1-4) were determined. Additionally, a large abdomen phantom of 48 cm diameter was examined with both protocols at equivalent tube current-time product in order to compare the dose efficiency. The patient groups differed significantly in dose (CTDIvol XXL 72.9±23 versus standard 16.7±11 mGy; intraindividual 64.1±20 versus 27.0±15 mGy). The image noise was generally somewhat higher in the XXL group but significantly lower in the intraindividual comparison (liver 24.2±14 HU versus 36.3±20 HU; p=0.03; fat 15.5±8 HU versus 26.2±12 HU; p=0.02). With ratings of 1.9±0.7 and 1.8±0.7 image quality did not differ significantly in general, whereas there was a clear difference in the intraindividual comparison (1.8±0.8 versus 3.0±1.2) and only the XXL protocol achieved diagnostic quality in all cases, while 43% of the examinations with the standard protocol were rated as non-diagnostic. The quantification of dose efficiency in the phantom scans yielded no significant difference between the protocols. Up to 150 kg body weight, CT can be performed with the standard technique at 120 kVp with tube current modulation. In larger patients diagnostic image quality can only be achieved reliably with the dual source XXL mode, although at considerably increased calculated dose. However, standard conversion factors yield false high values so that the estimation of the biologically relevant equivalent dose is very difficult. (orig.)