Johannes Uhlig1, Uwe Fischer2, Lorenz Biggemann3, Joachim Lotz3,4, Susanne Wienbeck3. 1. Institute for Diagnostic and Interventional Radiology, University Medical Center Göttingen, Robert-Koch-Street 40, 37075, Göttingen, Germany. johannes.uhlig@med.uni-goettingen.de. 2. Diagnostic Breast Center Göttingen, Göttingen, Germany. 3. Institute for Diagnostic and Interventional Radiology, University Medical Center Göttingen, Robert-Koch-Street 40, 37075, Göttingen, Germany. 4. German Cardiovascular Research Centre, Partnersite Goettingen, Göttingen, Germany.
Abstract
OBJECTIVES: To evaluate whether post-contrast cone-beam breast CT (CBBCT) alone is comparable to the current standard of combined pre- and post-contrast CBBCT regarding diagnostic accuracy and superior regarding radiation exposure. MATERIAL AND METHODS: This study included 49 women (61 breasts) with median age 57.9 years and BI-RADS 4/5 lesions diagnosed on mammography/ultrasound in density type c/d breasts. Two radiologists rated post-contrast CBBCT and pre- and post-contrast CBBCT with subtraction images on the BI-RADS scale separately for calculation of inter- and intra-observer agreement and in consensus for diagnostic accuracy assessment. Sensitivity, specificity, and area under the curve (AUC) were compared via McNemar test and DeLong method, respectively. Subtraction imaging misregistration were measured from 1 (no artifacts) to 4 (artifacts with width > 4 mm). RESULTS: A total of 100 lesion (51 malignant; 6 high risk; 43 benign) were included. AUC, sensitivity, and specificity showed no significant differences comparing post-contrast CBBCT alone versus pre- and post-contrast CBBCT (AUC 0.84 vs. 0.83, p = 0.643; sensitivity 0.89 vs. 0.85, p = 0.158; specificity 0.73 vs. 0.76, p = 0.655). Inter- and intra-observer agreement was excellent (intra-class correlation coefficient ICC = 0.76, ICC = 0.83, respectively). Radiation dose was significantly lower for post-contrast CBBCT alone versus pre- and post-contrast CBBCT (median average glandular radiation dose 5.9 mGy vs. 11.7 mGy, p < 0.001). High-degree misregistrations were evident in the majority of subtraction images (level 1/2/3/4 16.9%/27.1%/16.9%/39%), in particular for bilateral exams (3.2%/29.2%/8.3%/58.3%). CONCLUSION: Diagnostic accuracy of post-contrast CBBCT alone is comparable to pre- and post-contrast CBBCT in type c/d breasts, while yielding a significant twofold radiation dose reduction. KEY POINTS: • The diagnostic accuracy of post-contrast CBBCT alone is comparable to dual acquisition of pre- and post-contrast CBBCT. • Acquisition of the post-contrast CBBCT scan alone reduces radiation exposure compared to pre- and post-contrast CBBCT, thus countering one of the main limitations of CBBCT. • High-degree misregistration artifacts limit the interpretation of subtraction images from pre- and post-contrast CBBCT studies.
OBJECTIVES: To evaluate whether post-contrast cone-beam breast CT (CBBCT) alone is comparable to the current standard of combined pre- and post-contrast CBBCT regarding diagnostic accuracy and superior regarding radiation exposure. MATERIAL AND METHODS: This study included 49 women (61 breasts) with median age 57.9 years and BI-RADS 4/5 lesions diagnosed on mammography/ultrasound in density type c/d breasts. Two radiologists rated post-contrast CBBCT and pre- and post-contrast CBBCT with subtraction images on the BI-RADS scale separately for calculation of inter- and intra-observer agreement and in consensus for diagnostic accuracy assessment. Sensitivity, specificity, and area under the curve (AUC) were compared via McNemar test and DeLong method, respectively. Subtraction imaging misregistration were measured from 1 (no artifacts) to 4 (artifacts with width > 4 mm). RESULTS: A total of 100 lesion (51 malignant; 6 high risk; 43 benign) were included. AUC, sensitivity, and specificity showed no significant differences comparing post-contrast CBBCT alone versus pre- and post-contrast CBBCT (AUC 0.84 vs. 0.83, p = 0.643; sensitivity 0.89 vs. 0.85, p = 0.158; specificity 0.73 vs. 0.76, p = 0.655). Inter- and intra-observer agreement was excellent (intra-class correlation coefficient ICC = 0.76, ICC = 0.83, respectively). Radiation dose was significantly lower for post-contrast CBBCT alone versus pre- and post-contrast CBBCT (median average glandular radiation dose 5.9 mGy vs. 11.7 mGy, p < 0.001). High-degree misregistrations were evident in the majority of subtraction images (level 1/2/3/4 16.9%/27.1%/16.9%/39%), in particular for bilateral exams (3.2%/29.2%/8.3%/58.3%). CONCLUSION: Diagnostic accuracy of post-contrast CBBCT alone is comparable to pre- and post-contrast CBBCT in type c/d breasts, while yielding a significant twofold radiation dose reduction. KEY POINTS: • The diagnostic accuracy of post-contrast CBBCT alone is comparable to dual acquisition of pre- and post-contrast CBBCT. • Acquisition of the post-contrast CBBCT scan alone reduces radiation exposure compared to pre- and post-contrast CBBCT, thus countering one of the main limitations of CBBCT. • High-degree misregistration artifacts limit the interpretation of subtraction images from pre- and post-contrast CBBCT studies.
Authors: Avice O'Connell; David L Conover; Yan Zhang; Posy Seifert; Wende Logan-Young; Chuen-Fu Linda Lin; Lawrence Sahler; Ruola Ning Journal: AJR Am J Roentgenol Date: 2010-08 Impact factor: 3.959
Authors: Nicolas D Prionas; Karen K Lindfors; Shonket Ray; Shih-Ying Huang; Laurel A Beckett; Wayne L Monsky; John M Boone Journal: Radiology Date: 2010-09 Impact factor: 11.105
Authors: Posy Seifert; David Conover; Yan Zhang; Renee Morgan; Andrea Arieno; Stamatia Destounis; Patricia Somerville; Philip F Murphy Journal: Breast J Date: 2014-06-17 Impact factor: 2.431
Authors: Karen K Lindfors; John M Boone; Thomas R Nelson; Kai Yang; Alexander L C Kwan; DeWitt F Miller Journal: Radiology Date: 2008-01-14 Impact factor: 11.105
Authors: Shadi Aminololama-Shakeri; Craig K Abbey; Peymon Gazi; Nicolas D Prionas; Anita Nosratieh; Chin-Shang Li; John M Boone; Karen K Lindfors Journal: Eur J Radiol Date: 2015-10-01 Impact factor: 3.528