RPS 212 - New insights in paediatric body imaging

Purpose:

As an alternative to chest x-ray and CT, which exposes patients to ionising radiation, this paper aims to investigate the utility of fast 3T sequences for the detection of pulmonary abnormalities in children with pneumonia.

Methods and materials:

47 children with clinically suspected pneumonia were prospectively included in this study. All children underwent thoracic MRI (3T) and PA chest x-ray. 15 patients had contrast-enhanced thorax CT or HRCT. MRI protocol included axial and coronal T2-weighted spectral presaturation with inversion recovery (SPIR) MultiVane XD, axial echo-planar imaging (EPI), and diffusion-weighted imaging (DWI) with respiratory gating. Kappa statistics, Cochran Q and McNemar tests were used to investigate the results.

Results:

There was a statistically substantial agreement between chest radiographs and MRIs in detecting consolidation/infiltration (k=0.64), peribronchial thickening (0.78), and bronchiectasis (k=1), a moderate agreement for detecting cavities (k=0.54) and pleural effusion (k=0.44), and a fair agreement for detecting empyema (k=0.32) and bilateral involvement of the lungs (k=0.23). MRI was superior to chest x-ray in detecting bilateral involvement (p=0.000), lymph nodes (p=0.000), pleural effusion (p=0.000), and empyema (p=0.003). MRI detected all consolidation or infiltration detected on CT imaging. The kappa test showed a moderate agreement between MRI and CT in detecting pleural effusion (k=0.48) and ground-glass opacity (k=0.44), and a substantial or almost perfect agreement for all other pathologies (k=0.61-0.99). No statistically significant difference was observed between MRI and MDCT for detecting pneumonia associated pathologies by the McNemar test.

Conclusion:

Thoracic 3T MRI is an accurate and effective technique to evaluate children with pneumonia. MRI detected more pathologies than chest x-ray and had similar results compared with thorax CT.

Limitations:

The timeline between imaging modalities was 0-3 days. Patients older than 6 years old were included to avoid giving sedation to patients.

Ethics committee approval

The study was approved by the Ondokuz Mayis University Ethical Committee on 30th of March, 2017.

Funding:

Ondokuz Mayis University BAP-project office.

Purpose:

To investigate the role of intravoxel incoherent motion (IVIM) MRI for the study of placental microvascular and microstructural impairment in intrauterine growth restriction (IUGR) compared to normal placentae.

Methods and materials:

63 singleton pregnancies (49 normal and 14 IUGR subjects) were enrolled. Advanced diffusion-weighted EPI with 10 b-values (0, 10, 30, 50, 75, 100, 150, 400, 700, and 1,000s/mm2) was performed at 1.5T. MR body diffusion toolbox prototype software was used to obtain perfusion fraction (f), pseudo-diffusion coefficient (D*), and diffusion coefficient (D) maps. For each subject, ROIs were manually placed in whole foetal and whole maternal placenta and in three parenchymal areas (umbilical (U), central (C), and peripheral (P)) in both the foetal and maternal side. Differences between f, D, and D* mean values and their correlation with gestational age (GA) were investigated in both the normal and IUGR group.

Results:

f was significantly lower (p=10^-9) in IUGR compared to the normal group in whole maternal and whole foetal ROIs, in U-, C-, and P- maternal and foetal ROIs, whereas D was higher in IUGR compared to normal pregnancies (p=0.01) in whole, U-, C-, and P- foetal ROIs. D* was significantly lower in IUGR than in the healthy group in the whole maternal side (p=0.02) and in the foetal P-ROI. D showed a significant negative correlation with GA in the whole placenta (p<0.0001) and in U-, C-, and P- maternal and foetal ROIs.

Conclusion:

According to our results, the IVIM model is able to quantify in vivo microvascular and microstructural impairment of placental parenchyma in IUGR compared to normal pregnancies, detecting dysfunctional perfusion and diffusion qualities in vivo. Placental IVIM-MRI is a useful tool in the microstructural analysis of placental insufficiency affecting IUGR pregnancies.

Limitations:

The heterogeneity of the IUGR group.

Ethics committee approval

The study was approved by the local ethics committee.

Funding:

No funding was received for this work.

Purpose:

Systemic to pulmonary collateral flow (SPCF) is a significant haemodynamic burden in univentricular heart (UVH) patients after a bidirectional Glenn operation and diminishes to some extent after a total cavopulmonary connection (TCPC). By cardiac magnetic resonance imaging (CMR), pulmonary flow (Qp) and pulmonary-to-systemic-blood-flow ratio (Qp/Qs) can be estimated from pulmonary arteries and pulmonary veins. Our objective was to define if these methods differ in operated UVH patients and if this difference correlates to the amount of SPCF.

Methods and materials:

A retrospective single-centre study included 82 consecutive CMR examinations of patients with UVH (pre-TCPC stage n=17 (21%), post-TCPC n=65 (79%)) was conducted. For each patient, phase-contrast velocity mapping derived Qp was calculated twice; as the sum of the pulmonary arteries flows (QpPA) and pulmonary vein flows (QpPV). Qp/Qs ratios were calculated based on both QpPA and QpPV. The relationship between delta Qp/Qs (calculated as [(Qp_PV/Qs)-(Qp_PA/Qs)]) and the SPCF was computed using the Pearson correlation coefficient. Minimal clinically important difference (MCID) in Qp/Qs ratios was estimated as ≥0.5, ROC curve, and cut-off analyses were performed.

Results:

In the pre-TCPC group, SPCF was significantly higher compared to the post-TCPC group (26%(10%-60%) and 14%(3%-37%), respectively, p<0.001). In both groups, QpPA/Qs was smaller or rarely equal to QpPV/Qs (p<0.001). Delta Qp/Qs and SPCF showed a strong positive correlation (r=0.85, p<0.001). ROC analysis suggests the cut-off for MCID ≥0.5 at about 20% of SPCF with a sensitivity of 80%, specificity of 84%, and excellent accuracy (AUC=0.92±0.04).

Conclusion:

SPCF violates Qp and Qp/Qs ratio measurements based only on PA, resulting in their underestimation. At any stage of UVH palliation, SPCF exceeding 20% of stroke volume decreases the antegrade pulmonary artery flow remarkably.

Limitations:

n/a

Ethics committee approval

n/a

Funding:

No funding was received for this work.

Purpose:

Early and reliable diagnosis is important for neonatal patients with respiratory diseases (RD) which are common in the neonatal intensive care unit (NICU). Lung ultrasound (LUS) on the contrary of chest x-ray (CXR) and computed tomography present increased use in everyday practice due to their safety and mobility (bedside method) as well as their ability to produce real-time continuous images.

Methods and materials:

99 neonates from NICU with a gestational age ranging from 25 weeks-40 weeks and respiratory distress were included in this study. All babies after clinical estimation (Apgar score, vital signs, and laboratory tests etc.) underwent CXR followed by LUS. The findings were compared between them.

Results:

Clinically 77 (78.8%) neonates suffered from respiratory distress syndrome (RDS), 3 (3%) revealed pneumothorax, and 14 (14.1%) full-term neonates presented with transient tachypnea of the newborn. 2 (2%) neonates with pleural fluids, 1 (1%) with meconium aspiration syndrome, 1 (1%) with a hernia, and 2 (2%) with right upper lobe collapse were diagnosed. The basis of our diagnosis was clinical decision. According to this, the severity of illness was estimated by chest x-ray as well as by LUS. CXR had the most sensitivity and specificity in pneumothorax and hernia (100%) while LUS had 66% and 100%, respectively. In RDS, CXR achieved 73.9% sensitivity and 82% specificity, and in LUS, had 99% sensitivity but only 64.1% specificity.

Conclusion:

CXR remains the method of choice in RD diagnosis. LUS is a useful tool in NICU. It is reliable, easy-to-use, convenient, non-invasive, and free of a radiation method with the potential to become a tool for bedside dynamic respiratory monitoring.

Limitations:

n/a

Ethics committee approval

n/a

Funding:

No funding was received for this work.

Purpose:

Chronic aspiration is a common childhood cause of recurrent lower respiratory tract infection. Diagnosis is made clinically and supported by videofluoroscopy (VF), but chronic aspiration is often suggested when right upper and bilateral lower lobe disease is seen on chest radiography or CT, despite limited evidence to support this. The objective of this study was to investigate CT findings in VF-proven aspiration.

Methods and materials:

Children under 5 years with VF and CT data were identified from a speech and language therapy referrals database. Aspiration was assessed as present/absent on VF. 9 CT features of aspiration (atelectasis, bronchial wall thickening, mosaic attenuation, mucus plugging, bronchiectasis, ground glass, consolidation, volume loss, and interstitial lung disease) were assessed as present/absent and scored for severity/extent on a 3-point scale.

Results:

61 patients were identified (44 aspiration, 17 no aspiration, mean age 1.5 vs 2.1 years respectively, p=0.22). There were more CT features in the aspiration group (mean 6 vs 4.8) with greater extent/severity scores (7.3 vs 5.0), but neither reached significance (p=0.24 and 0.46). The presence of atelectasis (in any lobe) suggested aspiration (p=0.04), with aspiration extent nearing significance (p=0.07). No other feature reached statistical significance. Most features demonstrated a right upper, right lower, and left lower lobe predominance in both cohorts.

Conclusion:

Our small cohort supports the use of atelectasis on CT as a prompt for the investigation of aspiration. The RUL, RLL, and LLL predominant distribution alone is not diagnostic and VF continues as the gold standard investigation for aspiration.

Limitations:

Our small cohort is based at a tertiary heart and lung centre; patients often have other comorbidities.

Ethics committee approval

Currently under submission.

Funding:

No funding was received for this work.

Purpose:

Laryngotracheal stenosis (LTS) is often successfully corrected with open airway surgery. However, the majority of patients remain with respiratory and vocal sequelae. A better understanding of these sequelae is needed to improve clinical care and surgical interventions. The aim of this study was to image the upper airway of paediatric patients with a history of open airway surgery for LTS using magnetic resonance imaging (MRI).

Methods and materials:

48 patients (age 14.4 (range 7.5-30.7) years) and 11 healthy volunteers (age 15.9 (range 8.2-28.8) years) were included, spirometry and static and dynamic upper airway MRI (30 minutes, 3T GE scanner, 6CH carotid coil, spatial resolution 0.5x0.5 (in-plane) x2 mm, temporal resolution: 240 ms) were conducted. MRI analysis included the assessment of post-operative anatomy and measurements of airway areas and diameters during static and dynamic (inspiration and phonation) acquisitions.

Results:

Good image quality was achieved for static and dynamic images (61.4% and 70.2%) without artefacts (70.2%). Excellent MRI visualisation showed vocal cord thickening in 80.9% of patients and a significant decrease in lumen area at the vocal cord level (22.0 (IQR 17.7-30.3) mm² vs 35.1 (IQR 21.2-54.7) mm², p=0.03) but not the cricoid (62.0 ± 27.3 mm² vs 66.2 ± 34.8 mm², p=0.68) compared to healthy volunteers. Furthermore, 53.2% of patients had an a-frame deformation of the trachea at the site of the previous tracheal cannula, showing even further lumen collapse during inspiration. 37 (78.7%) patients showed signs of possible fibrosis/chronic oedema, mostly located at the arytenoids or vocal cords.

Conclusion:

Static and dynamic MRI is suitable for the extensive evaluation of the upper airway in post LTS children.

Limitations:

n/a

Ethics committee approval

Local medical ethics committee approval (MEC-2018-013) and written informed consent obtained.

Funding:

Funding provided by the ‘Vrienden van het Sophia Foundation’.

Purpose:

To evaluate acoustic radiation force impulse (ARFI) in assessing liver cirrhosis preoperatively in infants with biliary atresia (BA).

Methods and materials:

A total of 138 infants with BA scheduled for Kasai surgery were enrolled in this retrospective cohort study. All infants underwent ARFI imaging to evaluate liver stiffness by measuring shear-wave speeds (SWSs) followed by a liver biopsy within 3 days after imaging. The liver fibrosis stages of specimens were classified using the Batts-Ludwig scoring system. The Spearman correlation coefficient value between the SWSs and fibrosis stage was calculated. Receiver operating characteristic (ROC) curves were used for evaluating the diagnostic performance of ARFI imaging in diagnosing liver cirrhosis (fibrosis stage F4).

Results:

The median age of infants at the day of imaging was 56d. The biopsy histopathology confirmed 27 infants with fibrosis stage F1, 42 infants with fibrosis stage F2, 55 infants with fibrosis stage F3, and 14 infants with fibrosis stage F4. The median SWS of infants with F1, F2, F3, and F4 was 1.43m/s, 1.61m/s, 1.95m/s, and 2.39m/s, respectively (P<0.001). A significant correlation was found between SWS values and the fibrosis stage with a Spearman correlation coefficient of 0.744 (P<0.001). The AUC of SWS values for diagnosing cirrhosis was 0.948 (95% CI: 0.896, 0.978). The best cut-off value for predicting cirrhosis was 2.06m/s, with the sensitivity of 92.9% and specificity of 91.1%.

Conclusion:

The SWS value measured by ARFI imaging showed excellent correlation with the liver fibrosis stage and might be an ideal non-invasive approach in predicting liver cirrhosis in infants with BA preoperatively.

Limitations:

The retrospective study design.

Ethics committee approval

Appoved by insitute ethics committee (XHEC-D-2015-160).

Funding:

No funding was received for this work.

Purpose:

To evaluate the feasibility of acoustic force impulse (ARFI) elastography of the liver and spleen in patients with Gaucher disease type 1 (GD1) and to assess correlations between organ stiffness and clinico-radiologic data, such as the Gaucher disease type 1 severity scoring system (GD-DS3).

Methods and materials:

We retrospectively evaluated the results of ARFI elastography as measures of liver and spleen stiffness in 57 patients with GD1. The feasibility of the method was assessed. Correlations between elastography data and clinical data related to the metabolic syndrome, laboratory tests, and GD1-related clinico-radiologic data (bone marrow burden score, GD-DS3) were assessed.

Results:

ARFI elastography provided reliable results (i.e. standard deviation <30% of the mean value between the measurements) in 50/57 patients. Significant liver fibrosis was present in 35/50 patients (70%). Liver stiffness correlated significantly with the GD-DS3 score (P=0.029). Spleen stiffness correlated significantly with age (P=0.021), body mass index (P=0.002), several laboratory parameters, and near-significantly with the GD-DS3 score (P=0.059).

Conclusion:

ARFI elastography is a useful tool for a more profound assessment of disease severity in patients with GD1, which adds relevant information to the standard clinical scores. Thus, elastography might allow for extended therapy monitoring, especially in patients with significant liver fibrosis. Spleen elastography showed promising results and its role should be further investigated.

Limitations:

The retrospective study design and heterogeneity of the study cohort. Therapy-related factors were not analysed.

Ethics committee approval

The study was approved by the local independent ethics committee.

Funding:

Sanofi-Genzyme.

Purpose:

To determine whether the values of the hepatic apparent diffusion coefficient (ADC) can differentiate biliary atresia (BA) from non-BA or be correlated with the grade of hepatic fibrosis in infants with cholestasis.

Methods and materials:

This retrospective cohort study included infants who received liver MRI examinations to evaluate cholestasis from July 2009-October 2017. Liver ADC, ADC ratio of liver/spleen, aspartate aminotransferase to platelet ratio index (APRI), and spleen size were compared between the BA and non-BA groups. The diagnostic performances of all parameters for significant fibrosis (F3-4) were obtained by receiver-operating characteristics (ROC) curve analysis.

Results:

A total of 227 infants (M:F=98:129; mean age=57±36.3), including 125 BA patients, were analysed. Liver ADC values were lower (p=0.012) and spleen size was larger (p<0.001) in the BA group compared to the non-BA group. There were 34 patients with F0, 15 with F1, 71 with F2, 36 with F3, and 12 with F4. There were significant correlations between fibrosis grades and all 4 parameters of APRI (τ=0.289), spleen size (τ=0.297), liver ADC (τ=‑0.206), and ADC ratio (τ=-0.249) (all, p<0.001). The cut-off values for significant fibrosis (F3-4) were 0.712 for APRI (area under the ROC curve [AUC], 0.712), 5.9 cm for spleen size (AUC, 0.701), 1.044×10^-3 mm/sec for liver ADC (AUC, 0.666), and 1.22 for ADC ratio (AUC, 0.647).

Conclusion:

Liver ADC values could be helpful in differentiating BA from non-BA and predicting significant hepatic fibrosis in infants with cholestasis.

Limitations:

A retrospective study. The lack of evaluation for additional effects on conventional US or MR-based diagnoses.

Ethics committee approval

n/a

Funding:

No funding was received for this work.

Purpose:

To establish the normal reference values of a hepatic artery US Doppler postoperative evaluation after liver transplant in the paediatric population and to associate acute vascular complications with these US Doppler parameters.

Methods and materials:

A retrospective study of US Doppler parameters of liver transplants in children from 2016-2019 in a tertiary-care paediatrics hospital was performed.

Patient age, liver underlying disease, the introduction of enteral feeding, and surgical technique were recorded.

Peak systolic velocity (psV) and resistive index (RI) at US Doppler studies were analysed every 24 hours during the first 5 days after liver transplantation.

Results:

40 children (19 girls and 21 boys; mean age 5.4 years) enrolled during 2016-2019. One patient required retransplantation.

Of the 41 liver transplantations performed, 21 were living donor and deceased split-liver grafts.

6 acute arterial complications occurred, 4 anastomotic stenosis, and 2 arterial thromboses.

There was a significant association between psV and RI and with arterial complications. Range interval psV was inferior in partial-liver compared to total-liver transplants and in paediatric transplants compared to adults. A value of psV ≤30 cm/s after transplantation correlated with a risk of acute vascular complication and it was frequently observed after the 3rd day. A reduction of RI usually happened 1-2 days after the psV decrease.

Conclusion:

Normal reference psV range is inferior in paediatric liver transplant compared to that described in adults.

psV and RI correlate with arterial complications in the postoperative period. psV ≤30 cm/s value can predict potential acute arterial complications before RI.

CT should be performed to confirm US findings when psV ≤30 cm/s.

Limitations:

A retrospective study.

Ethics committee approval

n/a

Funding:

No funding was received for this work.

Purpose:

To compare the image quality of abdominal split-filter dual-energy CT (SF-DECT) to single-energy CT (SECT) and optimise the radiation dose using the paediatric phantom.

Methods and materials:

The abdomen regions of an anthropomorphic phantom representing a 5-year-old patient were scanned using SECT and SF-DECT retrospectively in different radiation dose groups. SF-DECT scans were reconstructed into composed images. Objective image quality, including CT values, image noise and contrast-to-noise (CNR) of liver and kidney, was measured and compared between SF-DECT and SECT in every radiation dose group. The results between SF-DECT and SECT were compared using the Mann–Whitney U test. After performing a Kruskal–Wallis test, objective image quality parameters were analysed among the different dose groups.

Results:

Image noise of SF-DECT was significantly lower than SECT in 3 mGy and 4 mGy (p＜0.001), and there was no difference in the 1 mGy and 2 mGy group (p=0.925 and 0.547, respectively). CNR of the liver and kidney of SF-DECT was significantly higher than SECT in all dose groups, except CNR of kidney in the 1 mGy group (p=0.862) and CNR of the liver in the 1 mGy and 2mGy group. The image noise decreased with the increase of dose in SF-DECT with the CNR increasing (p＜0.001). Furthermore, there were no differences between 3 mGy and 4 mGy groups for both the image noise and CNR.

Conclusion:

The use of split-filter dual-energy in paediatric abdominal CT could provide comparable, even superior image quality at different levels of radiation dose compared to standard SECT.

Limitations:

This is a phantom study that does not fully represent the condition of real children. We do not implement the subjective score of image quality.

Ethics committee approval

n/a

Funding:

No funding was received for this work.