Ali Pourvaziri1, Anushri Parakh1, Amirkasra Mojtahed1, Avinash Kambadakone1, Dushyant Vasudeo Sahani2. 1. Department of Radiology, Division of Abdominal Imaging and Intervention, Massachusetts General Hospital, Harvard Medical School, 55 Fruit St, White 270, Boston, MA, 02114, USA. 2. Department of Radiology, University of Washington, Box 357115, Seattle, WA, 98195-7115, USA. dsahani@uw.edu.
Abstract
PURPOSE: To compare the effect of dual-energy CT (DECT) material density datasets on diagnostic performance, readers' confidence, and interpretation time for renal lesion detection and characterization in comparison to subtraction CT (SCT). MATERIAL AND METHODS: One hundred fourteen patients (69/45 = M/F, mean age = 67 years) who underwent contrast-enhanced DECT between January 2015 and February 2018 for suspected renal mass were included retrospectively. For each patient, three radiologists assessed three image datasets: group A, material density iodine (MDI) + material density water (MDW); group B, SCT only; and group C, SCT + true unenhanced phase + virtual monochromatic images at 65 keV. Readers evaluated image quality (4-point scale), the number of lesions, and likely diagnosis. Reading times were recorded. Quantitatively, iodine concentration (IC from MDI) and delta Hounsfield units (ΔHU) for all lesions were measured. Diagnostic accuracy was compared using the area under the receiver operating characteristic curve (AUC). Image quality and interpretation time were compared with Kruskal-Wallis and t tests. RESULTS: Study cohort (230 lesions; mean size = 23.63 mm (5-116 mm)) consisted of 60 enhancing, 158 non-enhancing, and 12 lipid-dominant angiomyolipoma lesions. Significantly higher image quality was demonstrated for MDI compared to SCT (mean score = 3.82 vs. 3; p < 0.05). Comparable diagnostic accuracy was observed for group A (AUC = 0.88) and group C (AUC = 0.87) and was higher compared to that for group B (AUC = 0.75). Group A was read faster than group C (41.49 s vs. 71.45 s per exam; p < 0.05). Both IC and ΔHU values had high accuracy (AUC = 0.97) for differentiating enhancing vs. non-enhancing lesions; however, IC enabled differentiation of clear cell renal cell carcinoma from other enhancing lesions with moderate accuracy (AUC = 0.73). CONCLUSION: MDI images increase readers' confidence for renal lesion detection and characterization while providing a more efficient radiologist workflow, irrespective of readers' experience. KEY POINTS: • Material density iodine (MDI) images enable faster interpretation due to high image quality and potentially reduced need for quantitation. • MDI images increase diagnostic confidence of readers, irrespective of radiologists' experience. • High accuracy with dual-energy CT (DECT) can potentially reduce healthcare costs by eliminating the need for additional investigations.
PURPOSE: To compare the effect of dual-energy CT (DECT) material density datasets on diagnostic performance, readers' confidence, and interpretation time for renal lesion detection and characterization in comparison to subtraction CT (SCT). MATERIAL AND METHODS: One hundred fourteen patients (69/45 = M/F, mean age = 67 years) who underwent contrast-enhanced DECT between January 2015 and February 2018 for suspected renal mass were included retrospectively. For each patient, three radiologists assessed three image datasets: group A, material density iodine (MDI) + material density water (MDW); group B, SCT only; and group C, SCT + true unenhanced phase + virtual monochromatic images at 65 keV. Readers evaluated image quality (4-point scale), the number of lesions, and likely diagnosis. Reading times were recorded. Quantitatively, iodine concentration (IC from MDI) and delta Hounsfield units (ΔHU) for all lesions were measured. Diagnostic accuracy was compared using the area under the receiver operating characteristic curve (AUC). Image quality and interpretation time were compared with Kruskal-Wallis and t tests. RESULTS: Study cohort (230 lesions; mean size = 23.63 mm (5-116 mm)) consisted of 60 enhancing, 158 non-enhancing, and 12 lipid-dominant angiomyolipoma lesions. Significantly higher image quality was demonstrated for MDI compared to SCT (mean score = 3.82 vs. 3; p < 0.05). Comparable diagnostic accuracy was observed for group A (AUC = 0.88) and group C (AUC = 0.87) and was higher compared to that for group B (AUC = 0.75). Group A was read faster than group C (41.49 s vs. 71.45 s per exam; p < 0.05). Both IC and ΔHU values had high accuracy (AUC = 0.97) for differentiating enhancing vs. non-enhancing lesions; however, IC enabled differentiation of clear cell renal cell carcinoma from other enhancing lesions with moderate accuracy (AUC = 0.73). CONCLUSION:MDI images increase readers' confidence for renal lesion detection and characterization while providing a more efficient radiologist workflow, irrespective of readers' experience. KEY POINTS: • Material density iodine (MDI) images enable faster interpretation due to high image quality and potentially reduced need for quantitation. • MDI images increase diagnostic confidence of readers, irrespective of radiologists' experience. • High accuracy with dual-energy CT (DECT) can potentially reduce healthcare costs by eliminating the need for additional investigations.
Authors: Fawaz Al Harbi; Leila Tabatabaeefar; Michael A Jewett; Anthony Finelli; Martin O'Malley; Mostafa Atri Journal: AJR Am J Roentgenol Date: 2016-03 Impact factor: 3.959
Authors: Achille Mileto; Rendon C Nelson; Ehsan Samei; Tracy A Jaffe; Erik K Paulson; Andrew Barina; Kingshuk Roy Choudhury; Joshua M Wilson; Daniele Marin Journal: Radiology Date: 2014-05-15 Impact factor: 11.105
Authors: Amir A Borhani; Matthew Kulzer; Negaur Iranpour; Anish Ghodadra; Mark Sparrow; Alessandro Furlan; Mitchell E Tublin Journal: Abdom Radiol (NY) Date: 2017-03
Authors: Jonathan R Dillman; Elaine M Caoili; Richard H Cohan; James H Ellis; Isaac R Francis; Matthew J Schipper Journal: Abdom Imaging Date: 2008 Nov-Dec