Contents lists available at ScienceDirect Clinical Imaging journal homepage: www.elsevier.com/locate/clinimag Body Imaging Comparison of image quality and focal lesion detection in abdominopelvic CT: Potential dose reduction using advanced modelled iterative reconstruction Seung Joon Choi a,1 , So Hyun Park a, ⁎ , Young Sup Shim a,1 , Jung Han Hwang a , Suyoung Park a , Seong Yong Pak b , Myung-Won You c , Seong Ho Park d a Department of Radiology, Gil Medical Center, Gachon University College of Medicine, 21, Namdong-daero 774beon-gil, Namdong-gu, Incheon 21565, Republic of Korea b Imaging and Computer Vision Division, Siemens Healthcare, Seoul, Republic of Korea c Department of Radiology, Kyung Hee University Hospital, Seoul, Republic of Korea d Department of Radiology, Research Institute of Radiology, University of Ulsan College of Medicine, Asan Medical Center, 88, Olympic-ro 43-gil, Songpa-gu, Seoul 05505, Republic of Korea ARTICLE INFO Keywords: Abdominopelvic cavity Computed tomography Radiation dose reduction Dual-source CT Model-based iterative reconstruction Iterative reconstruction technique ABSTRACT Purpose: To evaluate radiation dose exposure, diagnostic performance, and image quality of low-dose and ul- tralow-dose abdominopelvic CT using the advanced modelled iterative reconstruction (ADMIRE) algorithm for focal lesion detection. Methods: One hundred thirty-nine consecutive patients underwent contrast-enhanced abdominopelvic CT using a third-generation dual-source scanner to obtain three data sets with the following tube loads: 33.3% (ultralow- dose CT), 66.7% (low-dose CT), and 100% (standard-dose CT). The diagnostic performances of standard-dose CT, low-dose CT, and ultralow-dose CT for focal lesion detection and characterization in organs of the abdo- minopelvic cavity were analyzed by two readers and compared with the reference standard. Image quality was measured subjectively and objectively. Results: Focal lesion detection showed 96.5–100% sensitivity and 97.7–100% accuracy in all representative organs on low-dose CT with acceptable image quality; it showed 87.4% sensitivity and 97.9% accuracy in the liver and 80.0% sensitivity and 88.2% accuracy in the rectal shelf on ultralow-dose CT with suboptimal image quality. Indeterminate lesions were significantly more common in the liver, pancreas, and kidneys on ultralow- dose CT than on low-dose CT. Enlarged lymph nodes showed 100% sensitivity and accuracy on ultralow-dose CT. Mean effective radiation doses of low-dose CT and ultralow-dose CT were 2.6 mSv and 1.3 mSv, respectively. Conclusions: The diagnostic performance of low-dose CT is similar to that of standard-dose abdominopelvic CT with acceptable image quality. Ultralow-dose CT cannot safely assess focal liver, pancreas, kidneys, and rectal shelf lesions but may be useful for the evaluation of enlarged lymph nodes. 1. Introduction With advances in early detection and effective treatments, cancer survival has improved in recent decades [1]. Contrast-enhanced CT is a useful imaging modality for diagnosis, staging, and treatment response monitoring in oncology. However, there are growing concerns re- garding lifetime risks of radiation exposure-induced cancer from CT [2]. Current CT systems have several dose-reduction strategies, including automatic tube current modulation, iterative reconstruction (IR) technique, and automatic exposure control [3,4]. IR reduces sta- tistical noise and enhances image quality, especially in low-dose CT scans [3–5]. Recently, IR techniques have been developed from statis- tical to model-based algorithms [6]. The advanced modelled iterative reconstruction algorithm (ADMIRE, Siemens, Erlangen, Germany) per- forms noise reduction in raw data and may enable greater dose re- duction with acceptable image quality [7,8]. A dual-source scanner can split or merge data from two tubes for https://doi.org/10.1016/j.clinimag.2020.01.017 Received 8 October 2019; Received in revised form 24 December 2019; Accepted 16 January 2020 ⁎ Corresponding author at: Department of Radiology, Gil Medical Center, Gachon University College of Medicine, 21, Namdong-daero 774beon-gil, Namdong-gu, Incheon 21565, Republic of Korea. E-mail address: nnoleeter@gilhospital.com (S.H. Park). 1 Contributed equally to this study. Clinical Imaging 62 (2020) 41–48 0899-7071/ © 2020 Elsevier Inc. All rights reserved. T