Sam J. Weisenthal BA, Abstract Co-Author: Nothing to Disclose

Vana M. Derderian BS, Abstract Co-Author: Nothing to Disclose

Les Roger Folio DO, MPH, Presenter: Nothing to Disclose

Ronald M. Summers MD, PhD, Abstract Co-Author: Royalties, iCAD, Inc Research funded, iCAD, Inc Stockholder, Johnson & Johnson Grant, Viatronix, Inc

Jianhua Yao PhD, Abstract Co-Author: Royalties, iCAD, Inc

BMI is commonly used as a surrogate for patient size to estimate radiation dose from DLP. We investi-gate the association between adiposity, scanned body volume, BMI, and DLP to seek a more accurate and accessible patient size metric to correlate with radiation dose. Because our method extracts directly from DICOM headers, is applied to each slice and is computationally inexpensive, it can serve as a shorthand organ and region-specific dose estimator.

Chest, abdomen and pelvis CT exams (n=73) were retrospectively analyzed by an automated fat and body volume measurement tool that used segmentation to calculate total body volume (TBV), total subcutane-ous fat (TSF), and total visceral fat (TVF) for every slice and exam. Slice-specific scan and dose metrics were obtained from DICOM headers by an in-house Radiation Exposure Extraction Engine (RE3). A mul-tivariable regression was used to associate TBV, TSF and TVF with scanner-modulated exposure at the slice level and with DLP at the exam level. A single variable regression was used to relate DLP to BMI data acquired from the RIS. The BMI and TBV, TSF, TVF regressions were compared with a paired t-test on their residuals.

DLP ranged from 352 to 1,961 mGy*cm; TBV from 46,578 to 155,747 cm3; and BMI from 17 to 46 kg/m2. At the exam level, the multivariable regression incorporating TBV, TSF, and TVF (R2=0.92) had higher correlation to DLP than did BMI (R2=0.74). The BMI residuals had a mean of 126.0±113.8 and the TBV, TSF, TVF had a mean of 75.8±58.1; a paired t-test indicated significant (p<0.001) difference be-tween the two. At the slice level, the correlation between TBV, TSF, TVF and radiation exposure held (R2 = 0.69).

We demonstrate TBV, TSF, and TVF as an accurate region-specific dose estimator that can be obtained from scan data alone (DICOM header and image data) does not rely on external data such as BMI.

Accurate and accessible region and size-specific dose estimations will help radiology departments produce meaningful patient-tailored dose data, optimizing internal and national quality benchmarks for dose tracking and reduction.

Weisenthal, S, Derderian, V, Folio, L, Summers, R, Yao, J, Body Volume and Adiposity as Metrics for Patient and Region-specific Dose Estimation: A Comparison with BMI.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14014684.html