Click here to


Are you sure ?

Yes, do it No, cancel

Size-Specific CT Dose Modulation for Low-Contrast Detectability

J James1*, V Rana2 , Y Liang3 (1) Indiana University, Imaging Sciences, Indianapolis, IN, (2) Houston Methodist Hospital, Houston, TX, (3) Indiana University Medical Center, Indianapolis, IN


(Wednesday, 7/17/2019) 10:30 AM - 11:00 AM

Room: Exhibit Hall | Forum 9

Purpose: Optimizing dose reduction while maintaining image-quality (IQ) is crucial in CT. Currently all commercial scanners have implemented dose-modulation paradigms, which adjusts tube output according to patient dimensions. “Dose-Right-Index (DRI)� algorithm on Philips’ scanners modulates dose dictated by user-choice of DRI with the assumption that given DRI maintains similar IQ for every patient size. However, qualitative observations from clinicians show degraded IQ as patient size increases. Aims of this study are to quantify limitations of a universal DRI approach for all patients and to establish size-dependent DRI-algorithm based on low-contrast-detectability (LCD).

Methods: Phantom experiments were performed using a Catphan504 phantom with additional Teflon slip-rings that simulated small(20-cm), normal(30-cm), medium-to-large(35-cm) and obese(45-cm) patient-sizes on Philips-scanner (Ingenuity-128) with iterative-reconstruction, iDose4=level 4. “Adult-Abdomen� protocol was used to scan at DRI-levels (16,18,20,22,24) corresponding to CTDI, 2.5 to 30-mGy. At each DRI, tube kVps were varied (80,100,120,140). ROIs were positioned at the center slice of the phantom image to measure noise. LCDs were evaluated within the low-contrast-module in the Catphan504.

Results: Experiments show that at a given DRI, scanner designed tube modulation does not provide a constant image quality (image noise nor LCD). From 20-cm to 45-cm, a 6-fold dose increase was observed with a doubling of noise and decreased LCD. For any given phantom size at any given dose, this power law applies, d*c=AD^α; ‘d’,‘c’: minimum detectable lesion size (mm) associated with the particular object contrast ‘c’ (Hounsfield units), D: dose; A, α: experimentally determined patient-size dependent coefficients. A similar, but not identical, relationship was established in GE and Siemens scanners. kVp switching showed that lower kVp is feasible in smaller patients (<30cm) to further reduce the dose while maintaining LCD.

Conclusion: The results have provided quantitative guidance on how to increase dose for a certain DRI level as-per patient size in Philips scanners.


Not Applicable / None Entered.


IM- CT: Radiation dosimetry & risk

Contact Email