Image quality and diagnostic accuracy of reduced-dose computed tomography enterography with model-based iterative reconstruction in pediatric Crohn’s disease patients

www.nature.com/scientificreports OPEN Image quality and diagnostic accuracy of reduced‑dose computed tomography enterography with model‑based iterative reconstruction in pediatric Crohn’s disease patients Yeoun Joo Lee1, Jae‑Yeon Hwang2*, Hwaseong Ryu2, Tae Un Kim2, Yong‑Woo Kim2, Jae Hong Park1, Ki Seok Choo2, Kyung Jin Nam2 & Jieun Roh2 This study assessed the image quality and diagnostic accuracy in determining disease activity of the terminal ileum of the reduced-dose computed tomography enterography using model-based iterative reconstruction in pediatric patients with Crohn’s disease (CD). Eighteen patients were prospectively enrolled and allocated to the standard-dose (SD) and reduced-dose (RD) computed tomography enterography (CTE) groups (n = 9 per group). Image quality, reader confidence in interpreting bowel findings, accuracy in determining active CD in the terminal ileum, and radiation dose were evaluated. Objective image quality did not show intergroup differences, except for image sharpness. Although reader confidence in detecting mural stratification, ulcer, and perienteric fat stranding of the RD-CTE were inferior to SD-CTE, RD-CTE correctly diagnosed active disease in all patients. The mean values of radiation dose metrics (SD-CTE vs. RD-CTE) were 4.3 versus 0.74 mGy, 6.1 versus 1.1 mGy, 211.9 versus 34.5 mGy∙cm, and 4.4 versus 0.7 mSv mGy∙cm for CTDIvol, size-specific dose estimation, doselength product, and effective dose, respectively. RD-CTE showed comparable diagnostic accuracy to SD-CTE in determining active disease of the terminal ileum in pediatric CD patients. However, image quality and reader confidence in detecting ulcer and perienteric fat stranding was compromised. Inflammatory bowel disease is characterized by chronic and debilitating inflammatory episodes of the gastrointestinal tract. Due to the increasing incidence of inflammatory bowel disease in the pediatric population1, imaging studies visualizing the bowel, including endoscopy, computed tomography enterography (CTE), and magnetic resonance enterography (MRE), have also been frequently performed in pediatric p atients2. Computed tomography (CT) is an excellent noninvasive tool for evaluating abdominal manifestations. Moreover, patients diagnosed with Crohn’s disease (CD) in childhood are twice as likely to have high cumulative radiation exposure compared to patients diagnosed at an older age even though reduced-dose (RD) CT protocol is currently being used for pediatric patients in many institutions3. A few technical parameters can be applied to minimize the radiation dose during a CT scan, including tube current modulation, peak kilovoltage (kVp) lowering, and the use of iterative reconstruction. Low kVp can increase contrast visualization, particularly in CT angiography, due to higher attenuation of iodine contrast media with lower tube voltage as photon energy decreases toward the K-edge energy of 33 k eV4 while achieving radiation dose r eduction4–6. However, low kVp CT images inherently increase quantum mottles because of the higher absorption of low-energy photons within the human body, particularly in larger patients4. Therefore, the low kVp CT scan may be suitable for pediatric patients because image noise and streak artifacts are reduced in small individuals7. 1 Department of Pediatrics, Pusan National University Children’s Hospital, College of Medicine, Pusan National University, Yangsan 50612, Republic of Korea. 2Department of Radiology, Research Institute for Convergence of Biomedical Science and Technology, Pusan National University Yangsan Hospital, College of Medicine, Pusan National University, Yangsan 50612, Republic of Korea. *email: Scientific Reports | (2022) 12:2147 | https://doi.org/10.1038/s41598-022-06246-z 1 Vol.:(0123456789) www.nature.com/scientificreports/ Model-based iterative reconstruction (MBIR; Veo, GE Healthcare, Milwaukee, WI, USA) enables drastic image noise reduction. MBIR is known to achieve a higher level of noise reduction of up to 60–80% of the standard filtered backprojection algorithm in adult patients with effective denoizing performance8,9. However, the main drawback of MBIR is associated with undesirable image features expressed as blotchy, pixelated, or plastic-like image texture compared with conventional filtered backprojection images. Several studies have evaluated the feasibility and diagnostic accuracy of reduced-dose CTE (RD-CTE) in adult and pediatric patients by applying low kVp and iterative reconstruction10–16. These studies achieved a significant dose reduction of approximately 30%–70% with acceptable image quality and diagnostic accuracy in detecting bowel abnormalities. However, it is believed that few prospective studies have evaluated the accuracy of RD-CTE in assessing small bowel abnormalities in pediatric CD patients. Therefore, the purpose of the current study was to assess the feasibility, image quality, and diagnostic accuracy in determining disease activity of the terminal ileum of the RD-CTE using low kVp and MBIR techniques in pediatric CD patients. Methods Patient enrolment. This prospective study was approved by the Institutional Review Board of Pusan National University Yangsan Hospital, Yangsan, Republic of Korea (IRB no: 04–2014-024). All methods were performed following the Declaration of Helsinki and HIPAA regulations. Written informed consent for CTE and ileocolonoscopy was obtained from each patient and their guardians. Inclusion criteria were as follows: patients between 9 and 18 years old, patients with clinically suspected or known CD, and patients clinically indicated to CTE and ileocolonoscopy to evaluate small bowel CD. The exclusion criteria were as follows: patients who had a contraindication for intravenous injection of iodinated contrast media or ileocolonoscopy, failure to evaluate the terminal ileum on ileocolonoscopy, patients who were intolerant to negative oral contrast media, patients who could not perform a 10-s breath-hold, and patients with high suspicion of bowel obstruction. Patients were grouped into the standard- or RD groups by a simple randomization method using computer-generated random numbers. Patients’ characteristics, clinical and laboratory findings (i.e., including pediatric CD activity index, C-reactive protein, and fecal calprotectin) were obtained by reviewing electrical medical records. CTE protocols and image reconstruction. Patients were required to fast for 6 h before the CTE. Negative oral contrast media (0.1% w/v barium solution; Easymark, TAEJOON PHARM Co. Ltd., Seoul, Korea) was administered 1 h before the CT scan. The total amount of negative oral contrast media were 1,000, 1,200, and 1,500 mL for patients weighing < 40, 40–59.9, and > 60 kg, respectively. In addition, 30%, 30%, 20%, and 20% of the total amount of negative oral contrast media was administered 60, 45, 30, and 15 min before the start of the CT examination, respectively. Moreover, n (...truncated)