Abstract Archives of the RSNA, 2014
SSM11-04
Distinguishing Adrenal Adenomas from Non-adenoma by Dual Energy CT Analysis with 320-row Area Detector CT: Comparison with Single Energy CT Analysis and Chemical Shift MR Imaging
Scientific Papers
Presented on December 3, 2014
Presented as part of SSM11: Genitourinary (Adrenal Masses)
Koji Takumi, Presenter: Nothing to Disclose
Yoshihiko Fukukura MD, PhD, Abstract Co-Author: Nothing to Disclose
Junichi Ideue, Abstract Co-Author: Nothing to Disclose
Tomokazu Umanodan, Abstract Co-Author: Nothing to Disclose
Tomohide Yoneyama, Abstract Co-Author: Nothing to Disclose
Hiroto Hakamada, Abstract Co-Author: Nothing to Disclose
Masanori Nakajo MD, Abstract Co-Author: Nothing to Disclose
Takashi Yoshiura MD, PhD, Abstract Co-Author: Nothing to Disclose
To determine whether dual energy CT analysis can help differentiate adrenal adenomas from non-adenomas.
The study population consisted of 64 patients with 76 adrenal lesions (25 lipid-rich adenomas,37 lipid-poor adenomas, and 14 non-adenomas) who underwent dual energy CT examination using a 320-row area detector CT. Unenhanced CT scan with 120 kVp followed by unenhanced dual energy CT scan with 80 and 135 kVp were performed in all patients. Chemical shift MR imaging with a 3.0-T unit was performed in 31 adrenal lesions. For each adrenal lesion, an ovoid region of interest as large as possible to cover the largest dimension was drawn. We evaluated the following 5 parameters: mean attenuation value on 120, 80 and 135 kVp (MAV120, MAV80, MAV135), difference of mean attenuation values between 80 and 135 kVp (DMAV135-80), and signal intensity (SI) index: [(SI on in-phase imaging–SI on opposed-phase imaging)/(SI on in-phase imaging)]×100%. Receiver-operating characteristic (ROC) curve was used to assess the ability of the 5 parameters to discriminate total or lipid-poor adenomas from non-adenomas. The sensitivity obtained at the threshold criteria of 100% specificity for diagnosing adrenal adenomas were compared between all CT parameters by using the McNemar’s test. Pearson’s correlation analysis was used to evaluate the correlation between DMAV135-80 and SI index.
For diagnosing total and lipid-poor adenomas, area under the ROC curve values were 0.869 and 0.780 for MAV120, 0.895 and 0.823 for MAV80, 0.844 and 0.738 for MAV135, 0.893 and 0.820 for DMAV135-80, and 0.969 and 0.950 for SI index, respectively. With the thresholds values of individual parameters to yield 100% specificity, the sensitivity for diagnosing total and lipid-poor adenomas were 61.3 and 35.1% for MAV120, 66.1, 43.2% for MAV80, 53.2 and 21.6% for MAV135, and 75.8 and 59.5% for DMAV135-80. The sensitivities for diagnosing total and lipid-poor adenoma were significantly higher with DMAV135-80 than with MAV120 (p=0.012,respectively) and MAV135 (p=0.001,respectively). DMAV135-80 values correlated well with SI index (r=0.650,p<0.001).
Dual energy CT analysis can help differentiate adrenal adenomas from non-adenomas.
Dual energy CT analysis can be used to detect lipid component, and improve diagnostic performance to differentiate adrenal adenomas from non-adenomas in comparison with single energy CT analysis.
Takumi, K,
Fukukura, Y,
Ideue, J,
Umanodan, T,
Yoneyama, T,
Hakamada, H,
Nakajo, M,
Yoshiura, T,
Distinguishing Adrenal Adenomas from Non-adenoma by Dual Energy CT Analysis with 320-row Area Detector CT: Comparison with Single Energy CT Analysis and Chemical Shift MR Imaging. Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL.
http://archive.rsna.org/2014/14010535.html