Prenatal diagnosis of hypoplastic left heart syndrome on ultrasound using artificial intelligence: How does performance compare to a current screening programme?

BACKGROUND: Artificial intelligence (AI) has the potential to improve prenatal detection of congenital heart disease. We analysed the performance of the current national screening programme in detecting hypoplastic left heart syndrome (HLHS) to compare with our own AI model. METHODS: Current screening programme performance was calculated from local and national sources. AI models were trained using four-chamber ultrasound views of the fetal heart, using a ResNet classifier. RESULTS: Estimated current fetal screening programme sensitivity and specificity for HLHS were 94.3% and 99.985%, respectively. Depending on calibration, AI models to detect HLHS were either highly sensitive (sensitivity 100%, specificity 94.0%) or highly specific (sensitivity 93.3%, specificity 100%). Our analysis suggests that our highly sensitive model would generate 45,134 screen positive results for a gain of 14 additional HLHS cases. Our highly specific model would be associated with two fewer detected HLHS cases, and 118 fewer false positives. CONCLUSION: If used independently, our AI model performance is slightly worse than the performance level of the current screening programme in detecting HLHS, and this performance is likely to deteriorate further when used prospectively. This demonstrates that collaboration between humans and AI will be key for effective future clinical use.

Cortical morphology at birth reflects spatiotemporal patterns of gene expression in the fetal human brain.

Interruption to gestation through preterm birth can significantly impact cortical development and have long-lasting adverse effects on neurodevelopmental outcome. We compared cortical morphology captured by high-resolution, multimodal magnetic resonance imaging (MRI) in n = 292 healthy newborn infants (mean age at birth = 39.9 weeks) with regional patterns of gene expression in the fetal cortex across gestation (n = 156 samples from 16 brains, aged 12 to 37 postconceptional weeks [pcw]). We tested the hypothesis that noninvasive measures of cortical structure at birth mirror areal differences in cortical gene expression across gestation, and in a cohort of n = 64 preterm infants (mean age at birth = 32.0 weeks), we tested whether cortical alterations observed after preterm birth were associated with altered gene expression in specific developmental cell populations. Neonatal cortical structure was aligned to differential patterns of cell-specific gene expression in the fetal cortex. Principal component analysis (PCA) of 6 measures of cortical morphology and microstructure showed that cortical regions were ordered along a principal axis, with primary cortex clearly separated from heteromodal cortex. This axis was correlated with estimated tissue maturity, indexed by differential expression of genes expressed by progenitor cells and neurons, and engaged in stem cell differentiation, neuron migration, and forebrain development. Preterm birth was associated with altered regional MRI metrics and patterns of differential gene expression in glial cell populations. The spatial patterning of gene expression in the developing cortex was thus mirrored by regional variation in cortical morphology and microstructure at term, and this was disrupted by preterm birth. This work provides a framework to link molecular mechanisms to noninvasive measures of cortical development in early life and highlights novel pathways to injury in neonatal populations at increased risk of neurodevelopmental disorder.

Complex diffusion-weighted image estimation via matrix recovery under general noise models.

We propose a patch-based singular value shrinkage method for diffusion magnetic resonance image estimation targeted at low signal to noise ratio and accelerated acquisitions. It operates on the complex data resulting from a sensitivity encoding reconstruction, where asymptotically optimal signal recovery guarantees can be attained by modeling the noise propagation in the reconstruction and subsequently simulating or calculating the limit singular value spectrum. Simple strategies are presented to deal with phase inconsistencies and optimize patch construction. The pertinence of our contributions is quantitatively validated on synthetic data, an in vivo adult example, and challenging neonatal and fetal cohorts. Our methodology is compared with related approaches, which generally operate on magnitude-only data and use data-based noise level estimation and singular value truncation. Visual examples are provided to illustrate effectiveness in generating denoised and debiased diffusion estimates with well preserved spatial and diffusion detail.

3-D Reconstruction in Canonical Co-Ordinate Space From Arbitrarily Oriented 2-D Images.

Limited capture range, and the requirement to provide high quality initialization for optimization-based 2-D/3-D image registration methods, can significantly degrade the performance of 3-D image reconstruction and motion compensation pipelines. Challenging clinical imaging scenarios, which contain significant subject motion, such as fetal in-utero imaging, complicate the 3-D image and volume reconstruction process. In this paper, we present a learning-based image registration method capable of predicting 3-D rigid transformations of arbitrarily oriented 2-D image slices, with respect to a learned canonical atlas co-ordinate system. Only image slice intensity information is used to perform registration and canonical alignment, no spatial transform initialization is required. To find image transformations, we utilize a convolutional neural network architecture to learn the regression function capable of mapping 2-D image slices to a 3-D canonical atlas space. We extensively evaluate the effectiveness of our approach quantitatively on simulated magnetic resonance imaging (MRI), fetal brain imagery with synthetic motion and further demonstrate qualitative results on real fetal MRI data where our method is integrated into a full reconstruction and motion compensation pipeline. Our learning based registration achieves an average spatial prediction error of 7 mm on simulated data and produces qualitatively improved reconstructions for heavily moving fetuses with gestational ages of approximately 20 weeks. Our model provides a general and computationally efficient solution to the 2-D/3-D registration initialization problem and is suitable for real-time scenarios.

An exploration of the potential utility of fetal cardiovascular MRI as an adjunct to fetal echocardiography.

OBJECTIVES: Fetal cardiovascular magnetic resonance imaging (MRI) offers a potential alternative to echocardiography, although in practice, its use has been limited. We sought to explore the need for additional imaging in a tertiary fetal cardiology unit and the usefulness of standard MRI sequences. METHODS: Cases where the diagnosis was not fully resolved using echocardiography were referred for MRI. Following a three-plane localiser, fetal movement was assessed with a balanced steady-state free precession (bSSFP) cine. Single-shot fast spin echo and bSSFP sequences were used for diagnostic imaging. RESULTS: Twenty-two fetal cardiac MRIs were performed over 12 months, at mean gestation of 32 weeks (26-38 weeks). The majority of referrals were for suspected vascular abnormalities (17/22), particularly involving the aortic arch (n = 10) and pulmonary vessels (n = 4). Single-shot fast spin echo sequences produced 'black-blood' images, useful for examining the extracardiac vasculature in these cases. BSSFP sequences were more useful for intracardiac structures. Real-time SSFP allowed for dynamic assessment of structures such as cardiac masses, with enhancement patterns also allowing for tissue characterisation in these cases. CONCLUSIONS: Fetal vascular abnormalities such as coarctation can be difficult to diagnose by using ultrasound. Fetal MRI may have an adjunctive role in the evaluation of the extracardiac vascular anatomy and tissue characterisation. © 2016 The Authors. Prenatal Diagnosis published by John Wiley & Sons, Ltd.

Magnetic resonance imaging of the newborn brain: automatic segmentation of brain images into 50 anatomical regions.

We studied methods for the automatic segmentation of neonatal and developing brain images into 50 anatomical regions, utilizing a new set of manually segmented magnetic resonance (MR) images from 5 term-born and 15 preterm infants imaged at term corrected age called ALBERTs. Two methods were compared: individual registrations with label propagation and fusion; and template based registration with propagation of a maximum probability neonatal ALBERT (MPNA). In both cases we evaluated the performance of different neonatal atlases and MPNA, and the approaches were compared with the manual segmentations by means of the Dice overlap coefficient. Dice values, averaged across regions, were 0.81±0.02 using label propagation and fusion for the preterm population, and 0.81±0.02 using the single registration of a MPNA for the term population. Segmentations of 36 further unsegmented target images of developing brains yielded visibly high-quality results. This registration approach allows the rapid construction of automatically labeled age-specific brain atlases for neonates and the developing brain.

Single bubble acoustic characterization and stability measurement of adherent microbubbles.

This article examines how the acoustic and stability characteristics of single lipid-shelled microbubbles (MBs) change as a result of adherence to a target surface. For individual adherent and non-adherent MBs, the backscattered echo from a narrowband 2-MHz, 90-kPa peak negative pressure interrogation pulse was obtained. These measurements were made in conjunction with an increasing amplitude broadband disruption pulse. It was found that, for the given driving frequency, adherence had little effect on the fundamental response of an MB. Examination of the second harmonic response indicated an increase of the resonance frequency for an adherent MB: resonance radius increasing of 0.3 ± 0.1 µm for an adherent MB. MB stability was seen to be closely related to MB resonance and gave further evidence of a change in the resonance frequency due to adherence.

Albumin coated microbubble optimization: custom fabrication and comprehensive characterization.

Gas microbubbles are used routinely to improve contrast in medical diagnostic imaging. The emerging fields of microbubble-enhanced quantitative imaging and microbubble-enhanced drug delivery have further enhanced the drive toward microbubble characterization and design techniques. The quest to improve efficiency, particularly in the field of drug delivery, presents a requirement to develop methods to manipulate microbubble properties to improve utility. This article presents an investigation in to the feasibility of influencing albumin shelled microbubble properties through the variation of albumin availability during fabrication. Microbubbles were fabricated from albumin suspensions of varying concentration before thorough physical and acoustic characterization. Microbubbles with shells fabricated from a 2% albumin suspension had a greater scattering to attenuation ratio (STAR) than 10% albumin preparations (4.4% and 2.2%, respectively) and approximately double the nonlinear STAR (from 0.7% to 1.5%). The 2% microbubbles also exhibited greater (up to 40%), more violent radial oscillations during high speed imaging than 5% and 10% preparations. The results show that microbubble characteristics can be simply manipulated in the lab and indicate that for a given application this may provide the opportunity to further enhance favorable characteristics.

Magnetic resonance imaging of the newborn brain: manual segmentation of labelled atlases in term-born and preterm infants.

Premature birth is a major and growing problem. Investigations into neuroanatomical correlates and consequences of preterm birth are hampered by complex neonatal brain anatomy and unavailability of atlases and protocols covering the whole brain. We developed delineation protocols for the manual segmentation of cerebral magnetic resonance (MR) images from newborn infants into 50 regions with comprehensive coverage of the brain. We then segmented MR scans from 15 infants born preterm at median 29, range 26-35, weeks postmenstrual age and scanned at term-corrected age, and five term-born infants born at median 41, range 39-45, weeks postmenstrual age. Total and regional brain volumes were estimated in each infant, and regional volumes expressed as a fraction of total brain volume. Total brain volumes were higher with greater age at birth and at time of scan, but once corrected for age at scan there was no difference between preterm and term infants. Fractional age-corrected regional volumes were bigger unilaterally in terms in middle and inferior temporal gyri, anterior temporal lobe, fusiform gyrus and posterior cingulate gyrus. Fractional age-corrected regional volumes were larger in preterms bilaterally in hippocampus, amygdala, thalamus and lateral ventricles, left superior temporal gyrus and right caudate nucleus. These differences were not significant after correcting for multiple hypothesis testing, but suggest subtle differences between preterms and term-borns accessible to regional analysis. Detailed illustrated protocols are made available in the Appendix.

Effect of albumin and dextrose concentration on ultrasound and microbubble mediated gene transfection in vivo.

Ultrasound and microbubble mediated gene transfection has great potential for site-selective, safe gene delivery. Albumin-based microbubbles have shown the greatest transfection efficiency but have not been optimised specifically for this purpose. Additionally, few studies have highlighted desirable properties for transfection specific microbubbles. In this article, microbubbles were made with 2% or 5% (w/v) albumin and 20% or 40% (w/v) dextrose solutions, yielding four distinct bubble types. These were acoustically characterised and their efficiency in transfecting a luciferase plasmid (pGL4.13) into female, CD1 mice myocardia was measured. For either albumin concentration, increasing the dextrose concentration increased scattering, attenuation and resistance to ultrasound, resulting in significantly increased transfection. A significant interaction was noted between albumin and dextrose; 2% albumin bubbles made with 20% dextrose showed the least transfection but the most transfection with 40% dextrose. This trend was seen for both nonlinear scattering and attenuation behaviour but not for resistance to ultrasound or total scatter. We have determined that the attenuation behaviour is an important microbubble characteristic for effective gene transfection using ultrasound. Microbubble behaviour can also be simply controlled by altering the initial ingredients used during manufacture.

The influence of gas saturation on microbubble stability.

Accurate acoustic characterisation is an essential component of any experimental investigation concerning the use and development of microbubble contrast agents. It is of increasing importance as applications such as therapy and molecular and quantitative imaging are investigated. Such characterisation is generally conducted in the laboratory in the form of bulk acoustic studies or optical observation of single bubbles using high speed photography in a water tank containing "out-gassed" water. The approach is widely used in acoustics to prevent inaccurate measurements being made due to the presence of gas bubbles settling on instrumentation, however, the term is often used to cover a range of water preparation techniques and the final gas content of the water is not usually stated. This technical note demonstrates the influence of gas content on the stability of microbubble contrast agents and concludes that characterisation should always be conducted in equilibrated, gas-saturated water to ensure accurate and repeatable measurements are made.

Disruption of intracardiac flow patterns in the newborn infant.

INTRODUCTION: Consistent patterns of rotational intracardiac flow have been demonstrated in the healthy adult human heart. Intracardiac rotational flow patterns are hypothesized to assist in the maintenance of kinetic energy of inflowing blood, augmenting cardiac function. Newborn cardiac function is known to be suboptimal secondary to decreased receptor number and sympathetic innervation, increased afterload, and increased reliance on atrial contraction to support ventricular filling. Patterns of intracardiac flow in the newborn have not previously been examined. RESULTS: Whereas 5 of the 13 infants studied showed significant evidence of rotational flow within the right atrium, 8 infants showed little or no rotational flow. Presence or absence of rotational flow was not related to gestational age, birth weight, postnatal age, atrial size, or image quality. Despite absence of intra-atrial rotational flow, atrioventricular valve flow into the left and right ventricles later in the cardiac cycle could be seen, suggesting that visualization techniques were adequate. DISCUSSION: While further study is required to assess its exact consequences on cardiac mechanics and energetics, disruption to intracardiac flow patterns could be another contributor to the multifactorial sequence that produces newborn circulatory failure. METHODS: We studied 13 newborn infants, using three-dimensional (3D) cardiac magnetic resonance phase-contrast imaging (spatial resolution 0.84 mm, temporal resolution 22.6 ms) performed without sedation/anesthesia.

Identification and characterisation of midbrain nuclei using optimised functional magnetic resonance imaging.

Localising activity in the human midbrain with conventional functional MRI (fMRI) is challenging because the midbrain nuclei are small and located in an area that is prone to physiological artefacts. Here we present a replicable and automated method to improve the detection and localisation of midbrain fMRI signals. We designed a visual fMRI task that was predicted would activate the superior colliculi (SC) bilaterally. A limited number of coronal slices were scanned, orientated along the long axis of the brainstem, whilst simultaneously recording cardiac and respiratory traces. A novel anatomical registration pathway was used to optimise the localisation of the small midbrain nuclei in stereotactic space. Two additional structural scans were used to improve registration between functional and structural T1-weighted images: an echo-planar image (EPI) that matched the functional data but had whole-brain coverage, and a whole-brain T2-weighted image. This pathway was compared to conventional registration pathways, and was shown to significantly improve midbrain registration. To reduce the physiological artefacts in the functional data, we estimated and removed structured noise using a modified version of a previously described physiological noise model (PNM). Whereas a conventional analysis revealed only unilateral SC activity, the PNM analysis revealed the predicted bilateral activity. We demonstrate that these methods improve the measurement of a biologically plausible fMRI signal. Moreover they could be used to investigate the function of other midbrain nuclei.

Influence of needle gauge on in vivo ultrasound and microbubble-mediated gene transfection.

Ultrasound and microbubble-mediated gene transfection are potential tools for safe, site-selective gene therapy. However, preclinical trials have demonstrated a low transfection efficiency that has hindered the progression of the technique to clinical application. In this paper it is shown that simple changes to the method of intravenous injection can lead to an increase in transfection efficiency when using 6-MHz diagnostic ultrasound and the ultrasound contrast agent, SonoVue. By using needles of progressively smaller gauge, i.e., larger internal diameter (ID), from 29 G (ID 0.184 mm) to 25 G (ID 0.31 mm), the transfection of a luciferase plasmid (pGL4.13) was significantly increased threefold in heart-targeted female CD1 mice. In vitro work indicated that the concentration and size distribution of SonoVue were affected by increasing needle gauge. These results suggest that the process of systemic delivery alters the bubble population and adversely affects transfection. This is exacerbated by using high-gauge needles. These findings demonstrate that the needle with the largest possible ID should be used for systemic delivery of microbubbles and genetic material.

Temperature-dependent differences in the nonlinear acoustic behavior of ultrasound contrast agents revealed by high-speed imaging and bulk acoustics.

Previous work by the authors has established that increasing the temperature of the suspending liquid from 20°C to body temperature has a significant impact on the bulk acoustic properties and stability of an ultrasound contrast agent suspension (SonoVue, Bracco Suisse SA, Manno, Lugano, Switzerland). In this paper the influence of temperature on the nonlinear behavior of microbubbles is investigated, because this is one of the most important parameters in the context of diagnostic imaging. High-speed imaging showed that raising the temperature significantly influences the dynamic behavior of individual microbubbles. At body temperature, microbubbles exhibit greater radial excursion and oscillate less spherically, with a greater incidence of jetting and gas expulsion, and therefore collapse, than they do at room temperature. Bulk acoustics revealed an associated increase in the harmonic content of the scattered signals. These findings emphasize the importance of conducting laboratory studies at body temperature if the results are to be interpreted for in vivo applications.

Functional cardiac MRI in preterm and term newborns.

OBJECTIVE: To use cardiac MRI techniques to assess ventricular function and systemic perfusion in preterm and term newborns, to compare techniques to echocardiographic methods, and to obtain initial reference data. DESIGN: Observational magnetic resonance and echocardiographic imaging study. SETTING: Neonatal Unit, Queen Charlotte's and Chelsea Hospital, London, UK. Patients 108 newborn infants with median birth weight 1627 (580-4140) g, gestation 32 (25-42) weeks. RESULTS: Mean (SD) flow volumes assessed by phase contrast (PC) imaging in 28 stable infants were left ventricular output (LVO) 222 (46), right ventricular output (RVO) 219 (47), superior vena cava (SVC) 95 (27) and descending aorta (DAo) 126 (32) ml/kg/min, with flow being higher at lower gestational age. Limits of agreement for repeated PC assessment of flow were LVO ±50.2, RVO ±55.5, SVC ±20.9 and DAo ±26.2 ml/kg/min. Mean (SD) LVO in 75 stable infants from three-dimensional models were 245 (47) ml/kg/min, with limits of agreement ±58.3 ml/kg/min. Limits of agreement for repeated echocardiographic assessment of LVO were ±108.9 ml/kg/min. CONCLUSIONS: Detailed magnetic resonance assessments of cardiac function and systemic perfusion are feasible in newborn infants, and provide more complete data with greater reproducibility than existing echocardiographic methods. Functional cardiac MRI could prove to be a useful research technique to study small numbers of newborn infants in specialist centres; providing insights into the pathophysiology of circulatory failure; acting as an outcome measure in clinical trials of inotropic intervention and so guiding clinical practice in the wider neonatal community.

Temperature dependent behavior of ultrasound contrast agents.

Recent interest in ultrasound contrast agents (UCAs) as tools for quantitative imaging and therapy has increased the need for accurate characterization. Laboratory investigations are frequently undertaken in a water bath at room temperature; however, implications for in vivo applications are not presented. Acoustic investigation of a bulk suspension of SonoVue (Bracco Research, Geneva, Switzerland) was made in a water bath at temperatures of 20-45 degrees C. UCA characteristics were significantly affected by temperature, particularly between 20 and 40 degrees C, leading to an increase in attenuation from 1.7-2.5 dB, respectively (p = 0.002) and a 2-dB increase in scattered signal over the same range (p = 0.05) at an insonation pressure of 100 kPa. Optical data supported the hypothesis that a temperature-mediated increase in diameter was the dominant cause, and revealed a decrease in bubble stability. In conclusion, measurements made at room temperature require careful interpretation with regard to behavior in vivo.

Longitudinal regional brain volume changes quantified in normal aging and Alzheimer's APP x PS1 mice using MRI.

In humans, mutations of amyloid precursor protein (APP) and presenilins (PS) 1 and 2 are associated with amyloid deposition, brain structural change and cognitive decline, like in Alzheimer's disease (AD). Mice expressing these proteins have illuminated neurodegenerative disease processes but, unlike in humans, quantitative imaging has been little used to systematically determine their effects, or those of normal aging, on brain structure in vivo. Accordingly, we investigated wildtype (WT) and TASTPM mice (expressing human APP(695(K595N, M596L)) x PS1(M146V)) longitudinally using MRI. Automated global and local image registration, allied to a standard digital atlas, provided pairwise segmentation of 13 brain regions. We found the mature mouse brain, unlike in humans, enlarges significantly from 6-14 months old (WT 3.8+/-1.7%, mean+/-SD, P<0.0001). Significant changes were also seen in other WT brain regions, providing an anatomical benchmark for comparing other mouse strains and models of brain disorder. In TASTPM, progressive amyloidosis and astrogliosis, detected immunohistochemically, reflected even larger whole brain changes (5.1+/-1.4%, P<0.0001, transgenexage interaction P=0.0311). Normalising regional volumes to whole brain measurements revealed significant, prolonged, WT-TASTPM volume differences, suggesting transgene effects establish at <6 months old of age in most regions. As in humans, gray matter-rich regions decline with age (e.g. thalamus, cerebral cortex and caudoputamen); ventricles and white matter (corpus callosum, corticospinal tract, fornix system) increase; in TASTPMs such trends often varied significantly from WT (especially hippocampus). The pervasive, age-related structural changes between WT and AD transgenic mice (and mouse and human) suggest subtle but fundamental species differences and AD transgene effects.

Numerically-simulated induced electric field and current density within a human model located close to a z-gradient coil.

PURPOSE: To simulate exposure (e.g., during interventional procedures) of a worker close to an operating MR scanner by calculating electric fields and current density within an anatomically realistic body model due to a z-gradient coil and to compare results with safety guidelines and European Directive 2004/40/EC. MATERIALS AND METHODS: Electric field and current density in an adult male model located at three positions within the range 0.19-0.44 m from the end of a generic z-gradient coil were calculated using the time-domain finite integration technique (FIT). Frequency scaling was used in which quasistatic conditions were assumed and results obtained at 1 MHz (assuming tissue conductivity values at 1 kHz) were scaled to 1 kHz. RESULTS: Current density (averaged over 1 cm(2)) in central nervous system (CNS) tissues up to 20.6 mA m(-2) and electric fields (averaged over 5 mm) up to 4.1 V m(-1) were predicted for a gradient of 10 mT m(-1) and slew rate of 10 T m(-1) second(-1). CONCLUSION: Compliance with 2004/40/EC, and with basic restriction values of Institute of Electrical and Electronics Engineers (IEEE) C95.6-2002, was predicted only at impracticably low gradients/slew rates in the ranges 4.9-9.1 mT m(-1)/4.9-9.1 T m(-1) second(-1) and 5-21 mT m(-1)/5-21 T m(-1) second(-1), respectively.

Early growth in brain volume is preserved in the majority of preterm infants.

OBJECTIVE: Preterm infants have reduced cerebral tissue volumes in adolescence. This study addresses the question: Is reduced global brain growth in the neonatal period inevitable after premature birth, or is it associated with specific medical risk factors? METHODS: Eighty-nine preterm infants at term equivalent age without focal parenchymal brain lesions were studied with 20 full-term control infants. Using a deformation-based morphometric approach, we transformed images to a reference anatomic space, and we used the transformations to calculate whole-brain volume and ventricular volume for each subject. Patterns of volume difference were correlated with clinical data. RESULTS: Cerebral volume is not reduced compared with term born control infants (p = 0.765). Supplemental oxygen requirement at 28 postnatal days is associated with lower cerebral tissue volume at term (p < 0.001), but there were no significant differences in cerebral volumes attributable to perinatal sepsis (p = 0.515) and quantitatively defined diffuse white matter injury (p = 0.183). As expected, the ventricular system is significantly larger in preterm infants at term equivalent age compared with term control infants (p < 0.001). INTERPRETATION: Cerebral volume is not reduced during intensive care for the majority of preterm infants, but prolonged supplemental oxygen dependence is a risk factor for early attenuation of global brain growth. The reduced cerebral tissue volume seen in adolescents born preterm does not appear to be an inevitable association of prematurity, but rather caused by either specific disease during intensive care or factors operating beyond the neonatal period.