Photon-counting computed tomography for paediatric congenital heart defects yields images of high diagnostic quality with low radiation doses at both 70 kV and 90 kV

Abstract

Background: Photon-counting computed tomography (PCCT) is a new clinical method that may show better diagnostic quality at lower radiation doses than conventional CT.

Objective: To investigate the diagnostic quality and radiation dose of paediatric cardiovascular PCCT for diagnosis of congenital heart defects at 70 kV and 90 kV.

Materials and methods: This retrospective assessment included clinical non-gated paediatric PCCT examinations for assessment of congenital heart defects. Radiation doses were recorded, and overall and specific diagnostic quality (1-4) were scored by four paediatric radiologists. Agreement, differences, and trends were assessed by percent rater agreement, intraclass correlation, Mann-Whitney tests, and Jonckheere-Terpstra tests.

Results: Seventy children with congenital heart defects were examined at 70 kV (n = 35; age 2 days-16 years; 63% boys) or 90 kV (n = 35; age 2 days-17 years; 51% boys). All observers gave a median score of 4 (high diagnostic quality) for both 70 kV and 90 kV, with no difference in median values between tube voltages (all P > 0.06). Agreement for overall scores was 66-94% for 70 kV and 60-77% for 90 kV. Agreement for specific scores was 80-97% for 70 kV and 83-89% for 90 kV. Size-dependent dose estimate was 0.68 mGy (0.25-2.02 mGy) for 70 kV and 1.10 mGy (0.58-2.71 mGy; P < 0.001) for 90 kV. Effective dose was 0.30 mSv (0.15-0.82 mSv) for 70 kV and 0.39 mSv (0.22-1.51 mSv; P = 0.01) for 90 kV.

Conclusion: Paediatric cardiovascular PCCT yields images for congenital heart defects of high diagnostic quality with low radiation dose at both 70 kV and 90 kV.

Keywords: Diagnosis; Heart defects, congenital; Paediatrics; Photon-counting computed tomography; Radiation dosage.

Fig. 1

Inclusion and exclusion flowchart. As this was an early installation of the clinical photon-counting computed tomography system, only 90 kV protocols were initially available. Therefore, consecutive patients were included at 90 kV until 70 kV protocols were made available by the vendor. Thereafter, as the clinical standard was to use the then available 70 kV protocols, the aim was to include as many examinations using 70 kV as already included using 90 kV. For 90 kV, one patient was excluded due to failed contrast agent timing and no possibility to reacquire the scan on the same occasion, and for 70 kV two patients were excluded due to corrupt dose data reported by the system

Fig. 2

Examples of contrast-enhanced photon-counting computed tomography (PCCT) images using the 70 kV protocol. a Coronal maximum intensity projection with a slice thickness of 6 mm in a 14-day-old boy with pulmonary atresia, who had received a modified Blalock-Thomas-Taussig shunt and later desaturated. The image shows an open shunt (arrowhead) between the right subclavian and pulmonary arteries with a slight narrowing in the proximal and distal anastomosis and a stenosis (arrow) in the right pulmonary artery. This case had a median overall diagnostic score of 4 and a median specific diagnostic score of 4 with a size-specific dose estimate of 0.32 mGy and an effective dose of 0.15 mSv. b Sagittal oblique image with a slice thickness of 0.6 mm in a 5-year-old boy with tetralogy of Fallot, corrected at four months of age with a pulmonary conduit. Echocardiography showed a suspected stenosis of the pulmonary conduit, which was clearly shown using PCCT (arrow). This case had a median overall diagnostic score of 4 and a median specific diagnostic score of 4 with a size-specific dose estimate of 1.16 mGy and an effective dose of 0.52 mSv. c Transaxial image with a slice thickness of 0.6 mm in a 3-month-old boy in whom echocardiography showed a suspected aortopulmonary window and anomalous origin of pulmonary artery branches. The PCCT image shows a large aortopulmonary window between the ascending aorta and the distal part of the pulmonary trunk (asterisk). This case had a median overall diagnostic score of 4 and a median specific diagnostic score of 4 with a size-specific dose estimate of 0.41 mGy and an effective dose of 0.30 mSv. d Coronal oblique image with a slice thickness of 0.6 mm in a 16-year-old girl born with pulmonary atresia, ventricular septal defect, and major aortopulmonary collateral arteries, unifocalised to a pulmonary conduit. Multiple stents were inserted due to stenosis of the pulmonary artery branches. The PCCT image shows a heavily calcified pulmonary conduit (arrows) and a pulmonary artery stent with intimal proliferation (arrowheads). Note the good visualisation of vessel lumens despite the presence of dense calcifications and a metal stent. This case had a median overall diagnostic score of 4 and a median specific diagnostic score of 4 with a size-specific dose estimate of 1.81 mGy and an effective dose of 0.80 mSv

Fig. 3

Examples of contrast-enhanced photon-counting computed tomography (PCCT) images using 90 kV protocols. a Sagittal oblique image with a slice thickness of 0.6 mm in an 8-day-old girl with suspected coarctation of the aorta by echocardiography. PCCT shows a hypoplastic arch (asterisk) with hypoplasia of the aortic isthmus and a large arterial duct (plus sign). This case had a median overall diagnostic score of 4 and a median specific diagnostic score of 4 with a size-specific dose estimate of 0.63 mGy and an effective dose of 0.22 mSv. b Transaxial image with a slice thickness of 0.6 mm in an 8-year-old girl with arterial vasculopathy, status post dilatation of the pulmonary arteries and the ascending aorta due to stenosis. Note the streak artefact from the relatively dense contrast medium (asterisk). Despite this, the pulmonary arteries and the ascending aorta were sufficiently visualised for diagnostic purposes. This case had a median overall diagnostic score of 3 and a median specific diagnostic score of 4 with a size-specific dose estimate of 1.09 mGy and an effective dose of 0.24 mSv. c, d Coronal (c) and transaxial (d) oblique images with a slice thickness of 0.6 mm depict the right (c; asterisk), and left (d; asterisk) pulmonary arteries in a 5-month-old prematurely born girl with idiopathic pulmonary arterial hypertension. The pulmonary arteries were well visualised without stenosis. This case had a median overall diagnostic score of 3 and a median specific diagnostic score of 3 with an effective dose of 0.30 mSv

Fig. 4

Computed tomography dose index (CTDI_vol) versus age and body surface area (BSA). For all patients below 1 year of age, CTDI_vol was less than 0.3 mGy for 70 kV, and less than 0.6 mGy for 90 kV. For patients between 1 year and 17 years of age, CTDI_vol increased with age for both 70 kV and 90 kV. It also increased with BSA for both 70 kV and 90 kV, but to a larger degree for 90 kV

Fig. 5

Size-specific dose estimate (SSDE) versus age and body surface area (BSA). For all patients below 1 year of age, SSDE was less than 0.65 mGy for 70 kV, and less than 1.4 mGy for 90 kV. For patients between 1 year and 17 years of age, SSDE was less than 2.05 mGy for 70 kV and less than 2.75 mGy for 90 kV. SSDE increased with BSA for both 70 kV and 90 kV, but to a larger degree for 90 kV

Fig. 6

Effective dose (E_eff) versus age and body surface area (BSA). For most patients below 1 year of age, E_eff was less than 0.5 mSv for 70kV, and less than 1 mSv for all patients between 1 year and 17 years of age. E_eff increased with BSA for both 70 kV and 90 kV, but to a larger degree for 90 kV