Dosimetric Feasibility of Cone-Beam CT in Pediatric Image-Guided Retrograde Gastrostomy Tube Insertions.

PURPOSE: Cone-beam computed tomography (CBCT) in interventional radiology allows volumetric imaging with open patient access. This work aimed to assess radiation dose metrics of CBCT in simulated image-guided retrograde gastrostomy (IGRG) tube insertions in pediatric anthropomorphic phantoms and to compare them to measured radiation dose metrics obtained using fluoroscopy during clinical IGRG tube insertions in children. METHODS: Radiation dose indices obtained from radiation dose structured reports of fluoroscopic IGRG tube insertions were retrospectively evaluated in a consecutive cohort of 30 children. Dose indices were fractionated into 3 clinical stages for each procedure (planning, insertion, and confirmation). These 3 stages in 30 patients (3 × 30 = 90 patient stages) were compared to dose indices measured from 4 CBCT acquisition protocols acquired in pediatric phantoms. RESULTS: The mean proportion of radiation dose during planning, insertion, and confirmation was 35%, 38% and 27%, with mean reference-point air kerma (range) measured to be 1.0 (0.02-6.0) mGy, 0.9 (0.03-4.1) mGy, and 0.7 (0.04-3.7) mGy, respectively. Cone-beam computed tomography dose varied greatly depending on technical parameters and protocol selection, ranging from 0.7 to 39.3 mGy. In 19% of patient stages, the most dose-sparing CBCT protocol evaluated on phantoms delivered less radiation than the radiation dose indices recorded from patient's fluoroscopy. CONCLUSIONS: From a dosimetric perspective, radiation delivered in CBCT can vary widely, yet can be appreciably low. With appropriate CBCT protocol selection, the radiation dose delivered may be sufficiently low to warrant consideration for use, if clinically needed during difficult IGRG tube insertions, and satisfy the interventionalist's benefit-risk assessment.

Three-dimensional rotational angiography in the assessment of vascular and airway compression in children after a cavopulmonary anastomosis.

The aim of the study was to examine the role of three-dimensional rotational angiography (3DRA) in assessing vascular and airway narrowing in children with a bidirectional cavopulmonary anastomosis (BCPA). The course of children with single ventricle physiology is often complicated by left pulmonary artery (LPA) and/or bronchial stenosis and may be related to aortic compression. 3DRA may be useful in evaluating this complex anatomy and possible mechanisms for the observed obstruction. Clinical data and imaging (2D angiography and 3DRA) of children with a BCPA were reviewed retrospectively. Measurements were taken at similar locations along the pulmonary arteries in both modalities and in the airways on 3DRA. Twenty-five children with a previous BCPA were assessed at mean age of 3.1 ± 2.0 years and weight of 13.6 ± 3.6 kg. Excellent correlation was found between 3DRA and 2D angiographic LPA measurements (r = 0.89, p < 0.0001). Twelve children had qualitative LPA stenosis on 3DRA, with a stenotic dimension of 6.6 ± 2.2 mm on 2D angiography and 6.8 ± 1.9 mm on 3DRA (r = 0.94, p < 0.0001). Ten cases with LPA stenosis also had bronchial stenosis (83 %). Qualitative airway assessment correlated with quantitative bronchial dimensions from 3DRA-derived tomographic images: Bronchial stenosis measured 4.4 ± 1.6 versus 5.9 ± 1.1 mm in those with a normal appearing bronchus (p = 0.009). Hybrid patients (initial palliation with bilateral pulmonary artery banding and arterial ductal stenting, n = 5) and all patients with a Damus-Kaye-Stansel (DKS) anastomosis (n = 9) were more likely to have LPA and left bronchial stenosis (OR 7.7, p = 0.04). 3DRA is a useful and accurate tool in assessment of LPA and airway narrowing after BCPA. Hybrid and DKS patients are more prone to LPA and bronchial stenosis, and 3DRA can provide insight into the mechanism.