Background: Many interventional radiological (IVR) procedures involve iodine contrast solution for planning, targeting, treatment, and evaluation. Measurement of the iodine distribution will lead us to quantitative, evidence-based IVR procedures. Our aim was to develop methods to quantify the iodine concentration for abdominal IVR procedures using phantom studies.

Methods: A C-arm CT scanner with a 19” flat-panel detector (Allura Xper FD20; Philips Medical Systems, Best, The Netherlands) and a 64-slice CT scanner (Sensation 64; Siemens Medical Solution, Erlangen, Germany) were used. As the limited detector size cannot image the entire abdomen, the resulting “truncation” artifact is inevitable in C-arm CT images for abdominal applications. Thus, we evaluated the following two aspects using an elliptic-cylindrical water bath phantom with inserts filled with iodine-water solution: (1) the effect of the truncated field-of-view by shifting the phantom (27-142 mm outside of the detector coverage) to simulate slim to very obese patients; and (2) the accuracy of iodine concentration with various mixture ratio (2.32-31.82 mgI/mL). In both settings, pixel values of C-arm CT images were converted to Hounsfield unit-like (HU_CCT) values by using the ratio of the pixel value with enhancement and the background. The iodine concentration was then calculated from the converted HU_CCT value using x-ray mass attenuation coefficient and mass density of iodine and water.

Results: From 27 mm to 142 mm of the truncation, the pixel values of C-arm CT images varied as much as 53%. However, the converted HU_CCT values varied less than 5%. With various iodine concentration (55-626 H.U. in CT images), the converted HU_CCT value of C-arm CT and pixel values of CT images showed very strong correlation (r2 = 0.99). The root-mean-squared errors of the estimated iodine concentration were 0.11 mgI/mL (-13 % to +9% error) with C-arm CT and 0.07 mgI/mL (-6% to +9% error) with CT, respectively.

Conclusions: The effect of truncation in abdominal procedures can be reduced from 53% to 5% by converting pixel values to the Hounsfield unit-like values (HU_CCT) using the proposed method. With the method, the iodine concentration can be quantitatively measured with an error of -13 % to +9 %.