P Chaudhary*, B Peng, T Lin, R DiTusa, H Hsu, B Navot, S Jambawalikar, K Hamacher, Columbia University Irving Medical Center, New York, NY

• 
  P Chaudhary

MO-I430-BReP-F1-1 (Monday, 7/11/2022) 4:30 PM - 5:30 PM [Eastern Time (GMT-4)]

Exhibit Hall | Forum 1

Purpose: ACR CT accreditation requires phantom images using the mid-liver technique for the average adult abdomen protocol. However, mid-liver is not a well-defined location and its mAs values can vary significantly within the patient population. This research aimed to produce a simple predictive measure that allows using the readily available CTDIvol value for adult abdomen scans to estimate a reasonable mid-liver mAs value.

Methods: Using the CT dose tracking software Radimetrics (Bayer Healthcare), patient exams using the most typical abdomen protocol, abdomen-pelvis and chest-abdomen-pelvis, were extracted for two CT scanners, GE Revolution HD and Siemens Biograph mCT, respectively. Scans were removed if the mean mAs approached their maximum value as the correlation between the patient thickness and mAs then was lost. For mid-liver location validation, the mAs values from two slices, the midpoint of all liver slices and the slice with the largest liver cross-section, were compared with the slice selected by an experienced radiologist. For both scanners, the sample size was increased until the changes in the mean of the mid-liver mAs, mean mAs, and CTDIvol were much smaller than the standard deviation (SD). The fitting between mid-liver mAs and CTDIvol were then performed using linear regression.

Results: For 10 reviewed exams, mAs variation among the three possible mid-liver slice locations was less than 5% of the mAs for the slice selected by the radiologist. For both scanners, 50 exams were more than sufficient to represent the CTDIvol distribution of abdomen exams. A good correlation between CTDIvol and mid-liver mAs was found.

Conclusion: For accreditation purposes, rather than scrolling through several patient scans to find the mid-liver mAs, it is possible to use the averaged CTDIvol from the dose reports of those scans and calculate the approximate mid-liver mAs value needed to satisfy the ACR requirement.

Keywords

Not Applicable / None Entered.

Taxonomy

Not Applicable / None Entered.

Contact Email