Outcome of partial agenesis of corpus callosum.

BACKGROUND: The diagnosis of corpus callosum anomalies by prenatal ultrasound has improved over the last decade because of improved imaging techniques, scanning skills, and the routine implementation of transvaginal neurosonography. OBJECTIVE: Our aim was to investigate all cases of incomplete agenesis of the corpus callosum and to report the sonographic characteristics, the associated anomalies, and the perinatal outcomes. STUDY DESIGN: We performed a retrospective analysis of cases from January 2007 to December 2017 with corpus callosum anomalies, either referred for a second opinion or derived from the prenatal ultrasound screening program in a single tertiary referral center. Cases with complete agenesis were excluded from the analysis. Standardized investigation included a detailed fetal ultrasound including neurosonogram, fetal karyotyping (standard karyotype or array comparative genomic hybridization) and fetal magnetic resonance imaging. The pregnancy outcome was collected, and pathologic investigation in case of termination of the pregnancy or fetal or neonatal loss was compared with the prenatal findings. The pregnancy and fetal or neonatal outcomes were reported. The neurologic assessment was conducted by a pediatric neurologist using the Bayley Scales of Infant Development-II and the standardized Child Development Inventory when the Bayley investigation was unavailable. RESULTS: Corpus callosum anomalies were diagnosed in 148 cases during the study period, 62 (41.9%) of which were excluded because of complete agenesis, and 86 fetuses had partial agenesis (58.1%). In 20 cases, partial agenesis (23.2%) was isolated, whereas 66 (76.7%) presented with different malformations among which 29 cases (43.9%) were only central nervous system lesions, 21 cases (31.8%) were non-central nervous system lesions, and 16 cases (24.3%) had a combination of central nervous system and non-central nervous system lesions. The mean gestational age at diagnosis for isolated and non-isolated cases was comparable (24.29 [standard deviation, 5.05] weeks and 24.71 [standard deviation, 5.35] weeks, respectively). Of the 86 pregnancies with partial agenesis, 46 patients opted for termination of the pregnancy. Neurologic follow-up data were available for 35 children. The overall neurologic outcome was normal in 21 of 35 children (60%); 3 of 35 (8.6%) showed mild impairment and 6 of 35 (17.1%) showed moderate impairment. The remaining 5 of 35 (14.3%) had severe impairment. The median duration of follow-up for the isolated form was 45.6 months (range, 36-52 months) and 73.3 months (range, 2-138 months) for the nonisolated form. CONCLUSION: Partial corpus callosum agenesis should be accurately investigated by neurosonography and fetal magnetic resonance imaging to describe its morphology and the associated anomalies. Genetic anomalies are frequently present in nonisolated cases. Efforts must be taken to improve ultrasound diagnosis of partial agenesis and to confirm its isolated nature to enhance parental counseling. Although 60% of children with prenatal diagnosis of isolated agenesis have a favorable prognosis later in life, they often have mild to severe disabilities including speech disorders at school age and behavior and motor deficit disorders that can emerge at a later age.

Longitudinal MRI in the context of in utero surgery for open spina bifida: A descriptive study.

INTRODUCTION: Fetal surgery for open spina bifida (OSB) requires comprehensive preoperative assessment using imaging for appropriate patient selection and to evaluate postoperative efficacy and complications. We explored patient access and conduct of fetal magnetic resonance imaging (MRI) for prenatal assessment of OSB patients eligible for fetal surgery. We compared imaging acquisition and reporting to the International Society of Ultrasound in Obstetrics and Gynecology MRI performance guidelines. MATERIAL AND METHODS: We surveyed access to fetal MRI for OSB in referring fetal medicine units (FMUs) in the UK and Ireland, and two NHS England specialist commissioned fetal surgery centers (FSCs) at University College London Hospital, and University Hospitals KU Leuven Belgium. To study MRI acquisition protocols, we retrospectively analyzed fetal MRI images before and after fetal surgery for OSB. RESULTS: MRI for fetal OSB was accessible with appropriate specialists available to supervise, perform, and report scans. The average time to arrange a fetal MRI appointment from request was 4 ± 3 days (range, 0-10), the average scan time available was 37 ± 16 min (range, 20-80 min), with 15 ± 11 min (range, 0-30 min) extra time to repeat sequences as required. Specific MRI acquisition protocols, and MRI reporting templates were available in only 32% and 18% of units, respectively. Satisfactory T2-weighted (T2W) brain imaging acquired in three orthogonal planes was achieved preoperatively in all centers, and 6 weeks postoperatively in 96% of FSCs and 78% of referring FMUs. However, for T2W spine image acquisition referring FMUs were less able to provide three orthogonal planes presurgery (98% FSC vs. 50% FMU, p < 0.001), and 6 weeks post-surgery (100% FSC vs. 48% FMU, p < 0.001). Other standard imaging recommendations such as T1-weighted (T1W), gradient echo (GE) or echoplanar fetal brain and spine imaging in one or two orthogonal planes were more likely available in FSCs compared to FMUs pre- and post-surgery (p < 0.001). CONCLUSIONS: There was timely access to supervised MRI for OSB fetal surgery assessment. However, the provision of images of the fetal brain and spine in sufficient orthogonal planes, which are required for determining eligibility and to determine the reversal of hindbrain herniation after fetal surgery, were less frequently acquired. Our evidence suggests the need for specific guidance in relation to fetal MRI for OSB. We propose an example guidance for MRI acquisition and reporting.

Gallbladder torsion: a rare cause of acute abdomen in a 12-month old child.

Background: Gallbladder torsion is a rare cause of an acute abdomen, predominantly occurring in elderly women and less frequently diagnosed in the pediatric population. The diagnosis is difficult and rarely made preoperatively. However, suspicion needs to be raised in children with acute onset of abdominal pain. Ultrasound can demonstrate different signs putting forward the diagnosis but findings are often non-specific, therefore clinical suspicion should prompt a laparoscopic exploration.Case presentation: We report a case of a 12-month old girl consulting with progressive abdominal discomfort and vomiting. Ultrasound revealed an enlarged gallbladder with thickening of the wall but without demonstrable color Doppler flow and a more horizontal orientation outside its normal anatomic fossa. Gallbladder torsion was suspected. Emergency laparoscopic exploration confirmed the diagnosis and a laparoscopic cholecystectomy was performed. The postoperative course was uneventful.Conclusions: Gallbladder torsion, although rare, should be included in the differential diagnosis of an acute abdomen in children. Early recognition is necessary for a favorable outcome. The diagnosis might be supported by ultrasound but remains difficult, which is why laparoscopic exploration should be considered when the diagnosis remains unclear.

Fetal-onset Alexander disease with radiological-neuropathological correlation.

Alexander disease is a leukodystrophy caused by mutations in the GFAP gene, primarily affecting the astrocytes. This report describes the prenatal and post-mortem neuroimaging findings in a case of genetically confirmed, fetal-onset Alexander disease with pathological correlation after termination of pregnancy. The additional value of fetal brain magnetic resonance imaging in the third trimester as a complementary evaluation tool to neurosonography is shown for suspected cases of fetal-onset Alexander disease. Diffuse signal abnormalities of the periventricular white matter in association with thickening of the fornix and optic chiasm can point towards the diagnosis. Furthermore, the presence of atypical imaging findings such as microcephaly and cortical folding abnormalities in this case broadens our understanding of the phenotypic variability of Alexander disease.

Fetal brain maceration score on postmortem magnetic resonance imaging vs. conventional autopsy.

BACKGROUND: Postmortem fetal magnetic resonance imaging (MRI) has been on the rise since it was proven to be a good alternative to conventional autopsy. Since the fetal brain is sensitive to postmortem changes, extensive tissue fixation is required for macroscopic and microscopic assessment. Estimation of brain maceration on MRI, before autopsy, may optimize histopathological resources. OBJECTIVE: The aim of the study is to develop an MRI-based postmortem fetal brain maceration score and to correlate it with brain maceration as assessed by autopsy. MATERIALS AND METHODS: This retrospective single-center study includes 79 fetuses who had postmortem MRI followed by autopsy. Maceration was scored on MRI on a numerical severity scale, based on our brain-specific maceration score and the whole-body score of Montaldo. Additionally, maceration was scored on histopathology with a semiquantitative severity scale. Both the brain-specific and the whole-body maceration imaging scores were correlated with the histopathological maceration score. Intra- and interobserver agreements were tested for the brain-specific maceration score. RESULTS: The proposed brain-specific maceration score correlates well with fetal brain maceration assessed by autopsy (τ = 0.690), compared to a poorer correlation of the whole-body method (τ = 0.452). The intra- and interobserver agreement was excellent (correlation coefficients of 0.943 and 0.864, respectively). CONCLUSION: We present a brain-specific postmortem MRI maceration score that correlates well with the degree of fetal brain maceration seen at histopathological exam. The score is reliably reproduced by different observers with different experience.

Current and future funding streams for paediatric postmortem imaging: European Society of Paediatric Radiology survey results.

BACKGROUND: Perinatal and childhood postmortem imaging has been accepted as a noninvasive alternative or adjunct to autopsy. However, the variation in funding models from institution to institution is a major factor prohibiting uniform provision of this service. OBJECTIVE: To describe current funding models employed in European and non-European institutions offering paediatric postmortem imaging services and to discuss the perceived barriers to future postmortem imaging service provision. MATERIALS AND METHODS: A web-based 16-question survey was distributed to members of the European Society of Paediatric Radiology (ESPR) and ESPR postmortem imaging task force over a 6-month period (March-August 2021). Survey questions related to the radiologic and autopsy services being offered and how each was funded within the respondent's institute. RESULTS: Eighteen individual responses were received (13/18, 72.2% from Europe). Only one-third of the institutions (6/18, 33.3%) have fully funded postmortem imaging services, with the remainder receiving partial (6/18, 33.3%) or no funding (5/18, 27.8%). Funding (full or partial) was more commonly available for forensic work (13/18, 72%), particularly where this was nationally provided. Where funding was not provided, the imaging and reporting costs were absorbed by the institute. CONCLUSION: Increased access is required for the expansion of postmortem imaging into routine clinical use. This can only be achieved with formal funding on a national level, potentially through health care commissioning and acknowledgement by health care policy makers and pathology services of the value the service provides following the death of a fetus or child. Funding should include the costs involved in training, equipment, reporting and image acquisition.

MRI Characteristics of Pediatric Renal Tumors: A SIOP-RTSG Radiology Panel Delphi Study.

BACKGROUND: The SIOP-Renal Tumor Study Group (RTSG) does not advocate invasive procedures to determine histology before the start of therapy. This may induce misdiagnosis-based treatment initiation, but only for a relatively small percentage of approximately 10% of non-Wilms tumors (non-WTs). MRI could be useful for reducing misdiagnosis, but there is no global consensus on differentiating characteristics. PURPOSE: To identify MRI characteristics that may be used for discrimination of newly diagnosed pediatric renal tumors. STUDY TYPE: Consensus process using a Delphi method. POPULATION: Not applicable. FIELD STRENGTH/SEQUENCE: Abdominal MRI including T1- and T2-weighted imaging, contrast-enhanced MRI, and diffusion-weighted imaging at 1.5 or 3 T. ASSESSMENT: Twenty-three radiologists from the SIOP-RTSG radiology panel with ≥5 years of experience in MRI of pediatric renal tumors and/or who had assessed ≥50 MRI scans of pediatric renal tumors in the past 5 years identified potentially discriminatory characteristics in the first questionnaire. These characteristics were scored in the subsequent second round, consisting of 5-point Likert scales, ranking- and multiple choice questions. STATISTICAL TESTS: The cut-off value for consensus and agreement among the majority was ≥75% and ≥60%, respectively, with a median of ≥4 on the Likert scale. RESULTS: Consensus on specific characteristics mainly concerned the discrimination between WTs and non-WTs, and WTs and nephrogenic rest(s) (NR)/nephroblastomatosis. The presence of bilateral lesions (75.0%) and NR/nephroblastomatosis (65.0%) were MRI characteristics indicated as specific for the diagnosis of a WT, and 91.3% of the participants agreed that MRI is useful to distinguish NR/nephroblastomatosis from WT. Furthermore, all participants agreed that age influenced their prediction in the discrimination of pediatric renal tumors. DATA CONCLUSION: Although the discrimination of pediatric renal tumors based on MRI remains challenging, this study identified some specific characteristics for tumor subtypes, based on the shared opinion of experts. These results may guide future validation studies and innovative efforts. LEVEL OF EVIDENCE: 3 Technical Efficacy Stage: 3.

What brain abnormalities can magnetic resonance imaging detect in foetal and early neonatal spina bifida: a systematic review.

PURPOSE: Open spina bifida (OSB) encompasses a wide spectrum of intracranial abnormalities. With foetal surgery as a new treatment option, robust intracranial imaging is important for comprehensive preoperative evaluation and prognostication. We aimed to determine the incidence of infratentorial and supratentorial findings detected by magnetic resonance imaging (MRI) alone and MRI compared to ultrasound. METHODS: Two systematic reviews comparing MRI to ultrasound and MRI alone were conducted on MEDLINE, EMBASE, and Cochrane databases identifying studies of foetal OSB from 2000 to 2020. Intracranial imaging findings were analysed at ≤ 26 or > 26 weeks gestation and neonates (≤ 28 days). Data was independently extracted by two reviewers and meta-analysis was performed where possible. RESULTS: Thirty-six studies reported brain abnormalities detected by MRI alone in patients who previously had an ultrasound. Callosal dysgenesis was identified in 4/29 cases (2 foetuses ≤ 26 weeks, 1 foetus under any gestation, and 1 neonate ≤ 28 days) (15.1%, CI:5.7-34.3%). Heterotopia was identified in 7/40 foetuses ≤ 26 weeks (19.8%, CI:7.7-42.2%), 9/36 foetuses > 26 weeks (25.3%, CI:13.7-41.9%), and 64/250 neonates ≤ 28 days (26.9%, CI:15.3-42.8%). Additional abnormalities included aberrant cortical folding and other Chiari II malformation findings such as lower cervicomedullary kink level, tectal beaking, and hypoplastic tentorium. Eight studies compared MRI directly to ultrasound, but due to reporting inconsistencies, it was not possible to meta-analyse. CONCLUSION: MRI is able to detect anomalies hitherto underestimated in foetal OSB which may be important for case selection. In view of increasing prenatal OSB surgery, further studies are required to assess developmental consequences of these findings.

Motion corrected fetal body magnetic resonance imaging provides reliable 3D lung volumes in normal and abnormal fetuses.

OBJECTIVES: To calculate 3D-segmented total lung volume (TLV) in fetuses with thoracic anomalies using deformable slice-to-volume registration (DSVR) with comparison to 2D-manual segmentation. To establish a normogram of TLV calculated by DSVR in healthy control fetuses. METHODS: A pilot study at a single regional fetal medicine referral centre included 16 magnetic resonance imaging (MRI) datasets of fetuses (22-32 weeks gestational age). Diagnosis was CDH (n = 6), CPAM (n = 2), and healthy controls (n = 8). Deformable slice-to-volume registration was used for reconstruction of 3D isotropic (0.85 mm) volumes of the fetal body followed by semi-automated lung segmentation. 3D TLV were compared to traditional 2D-based volumetry. Abnormal cases referenced to a normogram produced from 100 normal fetuses whose TLV was calculated by DSVR only. RESULTS: Deformable slice-to-volume registration-derived TLV values have high correlation with the 2D-based measurements but with a consistently lower volume; bias -1.44 cm3 [95% limits: -2.6 to -0.3] with improved resolution to exclude hilar structures even in cases of motion corruption or very low lung volumes. CONCLUSIONS: Deformable slice-to-volume registration for fetal lung MRI aids analysis of motion corrupted scans and does not suffer from the interpolation error inherent to 2D-segmentation. It increases information content of acquired data in terms of visualising organs in 3D space and quantification of volumes, which may improve counselling and surgical planning.

Antenatal management of congenital diaphragmatic hernia: What's next ?

Congenital diaphragmatic hernia can be diagnosed in the prenatal period and its severity can be measured by fetal imaging. There is now level I evidence that, in selected cases, Fetoscopic Endoluminal Tracheal Occlusion with a balloon increases survival to discharge from the neonatal unit as well as the risk for prematurity. Both effects are dependent on the time point of tracheal occlusion. Fetoscopic Endoluminal Tracheal Occlusion may also lead to iatrogenic death when the balloon cannot be timely retrieved. The implementation of the findings from our clinical studies, may also vary based on local conditions. These may be different in terms of available skill set, access to fetal therapy, as well as outcome based on local neonatal management. We encourage prior benchmarking of local outcomes with optimal postnatal management, based on large enough numbers and using identical criteria as in the recent trials. We propose to work further on prenatal prediction methods, and the improvement of fetal intervention. In this manuscript, we describe a research agenda from a fetal medicine perspective. This research should be in parallel with innovation in neonatal and pediatric (surgical) management of this condition.

Pre- and postnatal brain magnetic resonance imaging in congenital cytomegalovirus infection: a case report and a review of the literature.

BACKGROUND: Congenital cytomegalovirus infection (cCMV) is the most common known viral cause of neurodevelopmental delay in children. The risk of severe cerebral abnormalities and neurological sequelae is greatest when the infection occurs during the first trimester of pregnancy. Pre- and postnatal imaging can provide additional information and may help in the prediction of early neurological outcome. CASE PRESENTATION: This report presents the case of a newborn with cCMV infection with diffuse parenchymal calcifications, white matter (WM) abnormalities and cerebellar hypoplasia on postnatal brain imaging after magnetic resonance imaging (MRI) and neurosonogram (NSG) at 30 weeks showing lenticulostriate vasculopathy, bilateral temporal cysts and normal gyration pattern according to the gestational age (GA). No calcifications were seen on prenatal imaging. CONCLUSION: cCMV infection can still evolve into severe brain damage after 30 weeks of GA. For this reason, a two-weekly follow-up by fetal NSG with a repeat in utero MRI (iuMRI) in the late third trimester is recommended in cases with signs of active infection.

The prevalence of brain lesions after in utero surgery for twin-to-twin transfusion syndrome on third-trimester MRI: a retrospective cohort study.

OBJECTIVE: Due to the increased risk of antenatal brain lesions, we offer a third-trimester magnetic resonance imaging (MRI) scan to all patients who underwent an in utero intervention for twin-twin transfusion syndrome (TTTS). However, the usefulness of such a policy has not been demonstrated yet. Therefore, we determined the prevalence of antenatal brain lesions detected on third-trimester MRI and the proportion of lesions detected exclusively on MRI. MATERIALS AND METHODS: We conducted a retrospective cohort study of monochorionic diamniotic twin pregnancies complicated by TTTS that underwent laser coagulation of the vascular anastomoses or fetal reduction by umbilical cord occlusion between 2010 and 2017. We reviewed the third-trimester MRI findings and compared those with the prenatal ultrasonography. RESULTS: Of the 141 patients treated with laser coagulation and 17 managed by cord occlusion, 112/141 (79%) and 15/17 (88%) patients reached 28 weeks. Of those, 69/112 (62%) and 11/15 (73%) underwent an MRI between 28 and 32 weeks. After laser coagulation, MRI detected an antenatal brain lesion in 6 of 69 pregnancies (9%) or in 6 of 125 fetuses (5%). In 4 cases (67%), the lesion was detected only on MRI. In the 11 patients treated with cord occlusion, no brain lesions were diagnosed. CONCLUSION: The prevalence of brain lesions detected by third-trimester MRI is higher compared to prenatal ultrasonography alone, making MRI a useful adjunct to detect antenatal brain lesions in twin pregnancies after in utero treatment for TTTS. KEY POINTS: • In utero interventions for twin-to-twin transfusion syndrome (TTTS) do not prevent the occurrence of antenatal brain lesions. • Fetal magnetic resonance imaging (MRI) has high accuracy in detecting anomalies of cortical development and can be a useful adjunct to ultrasonography in diagnosing certain brain abnormalities. • After laser coagulation of the anastomoses for TTTS, third-trimester MRI diagnosed a brain lesion that was not detected earlier on ultrasound scan in 6% of pregnancies.

Cortical spectral matching and shape and volume analysis of the fetal brain pre- and post-fetal surgery for spina bifida: a retrospective study.

PURPOSE: A retrospective study was performed to study the effect of fetal surgery on brain development measured by MRI in fetuses with myelomeningocele (MMC). METHODS: MRI scans of 12 MMC fetuses before and after surgery were compared to 24 age-matched controls without central nervous system abnormalities. An automated super-resolution reconstruction technique generated isotropic brain volumes to mitigate 2D MRI fetal motion artefact. Unmyelinated white matter, cerebellum and ventricles were automatically segmented, and cerebral volume, shape and cortical folding were thereafter quantified. Biometric measures were calculated for cerebellar herniation level (CHL), clivus-supraocciput angle (CSO), transverse cerebellar diameter (TCD) and ventricular width (VW). Shape index (SI), a mathematical marker of gyrification, was derived. We compared cerebral volume, surface area and SI before and after MMC fetal surgery versus controls. We additionally identified any relationship between these outcomes and biometric measurements. RESULTS: MMC ventricular volume/week (mm3/week) increased after fetal surgery (median: 3699, interquartile range (IQR): 1651-5395) compared to controls (median: 648, IQR: 371-896); P = 0.015. The MMC SI is higher pre-operatively in all cerebral lobes in comparison to that in controls. Change in SI/week in MMC fetuses was higher in the left temporal lobe (median: 0.039, IQR: 0.021-0.054), left parietal lobe (median: 0.032, IQR: 0.023-0.039) and right occipital lobe (median: 0.027, IQR: 0.019-0.040) versus controls (P = 0.002 to 0.005). Ventricular volume (mm3) and VW (mm) (r = 0.64), cerebellar volume and TCD (r = 0.56) were moderately correlated. CONCLUSIONS: Following fetal myelomeningocele repair, brain volume, shape and SI were significantly different from normal in most cerebral layers. Morphological brain changes after fetal surgery are not limited to hindbrain herniation reversal. These findings may have neurocognitive outcome implications and require further evaluation.

Current state of perinatal postmortem magnetic resonance imaging: European Society of Paediatric Radiology questionnaire-based survey and recommendations.

BACKGROUND: Postmortem magnetic resonance imaging (MRI) in perinatal and childhood deaths is increasingly used as a noninvasive adjunct or alternative to autopsy. Imaging protocols vary between centres and consensus guidelines do not exist. OBJECTIVE: Our aim was to develop practical, standardised recommendations for perinatal postmortem MRI. MATERIALS AND METHODS: Recommendations were based on the results of two surveys regarding local postmortem MRI practices sent electronically to all 14 members of the European Society of Paediatric Radiology (ESPR) Postmortem Imaging Task Force and 17 members of the International Society of Forensic Radiology and Imaging Task Force (25 different centres). RESULTS: Overall, 11/14 (78.6%) respondents from different institutions perform postmortem MRI. All of these centres perform postmortem MRI for perinatal and neonatal deaths, but only 6/11 (54.5%) perform imaging in older children. CONCLUSION: We propose a clinical standard for postmortem MRI sequences plus optional sequences for neuroimaging and cardiac anatomy depending on available scanning time and referral indications.

An automated framework for localization, segmentation and super-resolution reconstruction of fetal brain MRI.

High-resolution volume reconstruction from multiple motion-corrupted stacks of 2D slices plays an increasing role for fetal brain Magnetic Resonance Imaging (MRI) studies. Currently existing reconstruction methods are time-consuming and often require user interactions to localize and extract the brain from several stacks of 2D slices. We propose a fully automatic framework for fetal brain reconstruction that consists of four stages: 1) fetal brain localization based on a coarse segmentation by a Convolutional Neural Network (CNN), 2) fine segmentation by another CNN trained with a multi-scale loss function, 3) novel, single-parameter outlier-robust super-resolution reconstruction, and 4) fast and automatic high-resolution visualization in standard anatomical space suitable for pathological brains. We validated our framework with images from fetuses with normal brains and with variable degrees of ventriculomegaly associated with open spina bifida, a congenital malformation affecting also the brain. Experiments show that each step of our proposed pipeline outperforms state-of-the-art methods in both segmentation and reconstruction comparisons including expert-reader quality assessments. The reconstruction results of our proposed method compare favorably with those obtained by manual, labor-intensive brain segmentation, which unlocks the potential use of automatic fetal brain reconstruction studies in clinical practice.

Fetal MRI for dummies: what the fetal medicine specialist should know about acquisitions and sequences.

Fetal MRI is an increasingly used tool in the field of prenatal diagnosis. While US remains the first line screening tool, as an adjuvant imaging tool, MRI has been proven to increase diagnostic accuracy and change patient counseling. Further, there are instances when US may not be sufficient for diagnosis. As a multidisciplinary field, it is important that every person involved in the referral, diagnosis, counseling and treatment of the patients is familiar with the basic principles, indications and findings of fetal MRI. The purpose of the current paper is to equip radiologists and non-radiologists with basic MRI principles and essential topics in patient preparation and provide illustrative examples of when fetal MRI may be used. This aims to aid the referring clinician in better selecting and improve patient counseling prior to arrival in the radiology department and, ultimately, patient care.

Aleatoric uncertainty estimation with test-time augmentation for medical image segmentation with convolutional neural networks.

Despite the state-of-the-art performance for medical image segmentation, deep convolutional neural networks (CNNs) have rarely provided uncertainty estimations regarding their segmentation outputs, e.g., model (epistemic) and image-based (aleatoric) uncertainties. In this work, we analyze these different types of uncertainties for CNN-based 2D and 3D medical image segmentation tasks at both pixel level and structure level. We additionally propose a test-time augmentation-based aleatoric uncertainty to analyze the effect of different transformations of the input image on the segmentation output. Test-time augmentation has been previously used to improve segmentation accuracy, yet not been formulated in a consistent mathematical framework. Hence, we also propose a theoretical formulation of test-time augmentation, where a distribution of the prediction is estimated by Monte Carlo simulation with prior distributions of parameters in an image acquisition model that involves image transformations and noise. We compare and combine our proposed aleatoric uncertainty with model uncertainty. Experiments with segmentation of fetal brains and brain tumors from 2D and 3D Magnetic Resonance Images (MRI) showed that 1) the test-time augmentation-based aleatoric uncertainty provides a better uncertainty estimation than calculating the test-time dropout-based model uncertainty alone and helps to reduce overconfident incorrect predictions, and 2) our test-time augmentation outperforms a single-prediction baseline and dropout-based multiple predictions.

Pericardio-Amniotic Shunting for Incomplete Pentalogy of Cantrell.

A 27-year-old woman, gravida 2, para 0, presented with an incomplete Pentalogy of Cantrell with an omphalocele, diaphragmatic hernia, and a pericardial defect at 32 weeks' gestation. A large pericardial effusion compressed the lungs and had led to a reduced lung growth with an observed-to-expected total lung volume of 28% as measured by MRI. The effusion disappeared completely after the insertion of a pericardio-amniotic shunt at 33 weeks. After birth, the newborn showed no signs of pulmonary hypoplasia and underwent a surgical correction of the defect. Protracted wound healing and a difficult withdrawal from opioids complicated the neonatal period. The child was discharged on postnatal day 105 in good condition. This case demonstrates that in case of Pentalogy of Cantrell with large pericardial effusion, the perinatal outcome might be improved by pericardio-amniotic shunting.

In vivo evidence by magnetic resonance volumetry of a gestational age dependent response to tracheal occlusion for congenital diaphragmatic hernia.

OBJECTIVE: We aimed to assess in vivo changes in lung and liver volumes in fetuses with isolated congenital diaphragmatic hernia, either expectantly managed or treated in utero. METHOD: This is a secondary analysis of prospectively collected data at two fetal therapy centers. We used archived magnetic resonance images of fetuses taken ≥7 days apart, creating paired observations in 20 expectantly managed cases, 41 with a second magnetic resonance prior to balloon reversal and 64 after balloon removal. We measured observed to expected total fetal lung volume (O/E TFLV) and liver-to-thoracic volume ratio. We calculated changes in volume as compared with the initial measurement and its rate as a function of gestational age (GA) at occlusion. RESULTS: The liver-to-thoracic volume ratio did not change in either group. In expectantly managed fetuses, O/E TFLV did not increase with gestation. In fetuses undergoing tracheal occlusion, the measured increase in volume was 2.6 times larger with balloon in place as compared with that after its removal. GA at tracheal occlusion was an independent predictor of the O/E TFLV. The net rate seems to initially increase and plateau at a maximum of 1.5% per week by 35 to 45 days after occlusion. CONCLUSIONS: Tracheal occlusion induces a net increase in volume, its magnitude essentially dependent on the GA at occlusion.

A Dempster-Shafer Approach to Trustworthy AI With Application to Fetal Brain MRI Segmentation.

Deep learning models for medical image segmentation can fail unexpectedly and spectacularly for pathological cases and images acquired at different centers than training images, with labeling errors that violate expert knowledge. Such errors undermine the trustworthiness of deep learning models for medical image segmentation. Mechanisms for detecting and correcting such failures are essential for safely translating this technology into clinics and are likely to be a requirement of future regulations on artificial intelligence (AI). In this work, we propose a trustworthy AI theoretical framework and a practical system that can augment any backbone AI system using a fallback method and a fail-safe mechanism based on Dempster-Shafer theory. Our approach relies on an actionable definition of trustworthy AI. Our method automatically discards the voxel-level labeling predicted by the backbone AI that violate expert knowledge and relies on a fallback for those voxels. We demonstrate the effectiveness of the proposed trustworthy AI approach on the largest reported annotated dataset of fetal MRI consisting of 540 manually annotated fetal brain 3D T2w MRIs from 13 centers. Our trustworthy AI method improves the robustness of four backbone AI models for fetal brain MRIs acquired across various centers and for fetuses with various brain abnormalities.