Objectives Artificial intelligence (AI) has shown promise in improving the performance of fetal ultrasound screening in detecting congenital heart disease (CHD). The effect of giving AI advice to human operators has not been studied in this context. Giving additional information about AI model workings, such as confidence scores for AI predictions, may be a way of further improving performance. Our aims were to investigate whether AI advice improved overall diagnostic accuracy (using a single CHD lesion as an exemplar), and to determine what, if any, additional information given to clinicians optimized the overall performance of the clinician-AI team. Methods An AI model was trained to classify a single fetal CHD lesion (atrioventricular septal defect (AVSD)), using a retrospective cohort of 121 130 cardiac four-chamber images extracted from 173 ultrasound scan videos (98 with normal hearts, 75 with AVSD); a ResNet50 model architecture was used. Temperature scaling of model prediction probability was performed on a validation set, and gradient-weighted class activation maps (grad-CAMs) produced. Ten clinicians (two consultant fetal cardiologists, three trainees in pediatric cardiology and five fetal cardiac sonographers) were recruited from a center of fetal cardiology to participate. Each participant was shown 2000 fetal four-chamber images in a random order (1000 normal and 1000 AVSD). The dataset comprised 500 images, each shown in four conditions: (1) image alone without AI output; (2) image with binary AI classification; (3) image with AI model confidence; and (4) image with grad-CAM image overlays. The clinicians were asked to classify each image as normal or AVSD. Results A total of 20 000 image classifications were recorded from 10 clinicians. The AI model alone achieved an accuracy of 0.798 (95% CI, 0.760-0.832), a sensitivity of 0.868 (95% CI, 0.834-0.902) and a specificity of 0.728 (95% CI, 0.702-0.754), and the clinicians without AI achieved an accuracy of 0.844 (95% CI, 0.834-0.854), a sensitivity of 0.827 (95% CI, 0.795-0.858) and a specificity of 0.861 (95% CI, 0.828-0.895). Showing a binary (normal or AVSD) AI model output resulted in significant improvement in accuracy to 0.865 (P < 0.001). This effect was seen in both experienced and less-experienced participants. Giving incorrect AI advice resulted in a significant deterioration in overall accuracy, from 0.761 to 0.693 (P < 0.001), which was driven by an increase in both Type-I and Type-II errors by the clinicians. This effect was worsened by showing model confidence (accuracy, 0.649; P < 0.001) or grad-CAM (accuracy, 0.644; P < 0.001). Conclusions AI has the potential to improve performance when used in collaboration with clinicians, even if the model performance does not reach expert level. Giving additional information about model workings such as model confidence and class activation map image overlays did not improve overall performance, and actually worsened performance for images for which the AI model was incorrect. (c) 2024 The Authors. Ultrasound in Obstetrics & Gynecology published by John Wiley & Sons Ltd on behalf of International Society of Ultrasound in Obstetrics and Gynecology.

Fetal cases with known genetic syndromes and GA matched controls were retrospectively selected. All had same day US and MRI (iFIND-2 project: REC:14/LO/1806). 3D MRI was obtained by combining multiple stacks of T2 weighted slices using slice to volume reconstruction. A semi-automated segmentation was performed and then visualised as a 3D model. A fetal sonographer/radiographer compared the 3D MRI to the 3D surface rendered US and the quality scored from suboptimal to excellent (1 to 5) or not achievable (0). 22q11.2 microdeletion syndrome and T21 cases (4 in total) and 4 GA matched controls were included (mean GA = 32+0 weeks). MRI reconstructions and 3D models were successfully achieved in all cases, quality range = 1 to 5. In 2 cases the 3D US of the fetal face was not achievable and US quality range = 3 to 5. The 3D MRI fetal craniofacial models demonstrated craniofacial surface anatomy, which were suitable for assessment of curved structures e.g. palpebral fissures, philtrum smoothness/length, malformed/malrotated ears and head shape. It is feasible to convert 2D MRI slices to 3D MRI volumes, which can then produce 3D fetal craniofacial models successfully in normal and high-risk cases, where US alone may be limited. Novel MRI post-processing methods will improve the quality and reproducibility of 3D MRI reconstructions, additionally an automated segmentation pipeline will increase the clinical utility. Imaging could be coupled with objective computer vision methods, (automated 3D facial analysis), to more accurately detect a dysmorphic phenotype in complex fetal cases. This approach can provide unique visualisation of craniofacial morphology in-utero thus improving diagnostic characterisation.

This paper proposes a multiresolution methodology to visualise and analyse big complex networks. The approach is useful for sensor placement in water distribution systems. Traditional approaches such as CPLEX for facility location in networked structures and eigenvector centrality measures benefit from being addressed at various hierarchical levels of coarseness.

La resiliencia hidráulica puede ser formulada como una medida de la capacidad de la red de distribución de agua para mantener un nivel de servicio mínimo bajo condiciones de fallo. Este artículo explora un modelo híbrido que une medidas de resiliencia hidráulica y de teoría de grafos. La propuesta es extender el concepto de distancia geodésica en las tuberías, teniendo en cuenta pérdidas de energía asociadas con el caudal. Nuevos algoritmos basados en caminatas aleatorias evalúan rutas hidráulicamente factibles e identifican diferentes niveles de resiliencia en nodos. Aquellos de menor valor en dicha evaluación son analizados en una segunda fase, considerando la disponibilidad y capacidad de sus rutas de abastecimiento. La criticidad de una tubería, medida a través de su impacto en la interrupción del suministro a los nodos de la red, también se analiza dentro de este novedoso marco con resultados fiables y computacionalmente eficientes.

La resiliencia hidráulica puede ser formulada como una medida de la capacidad de la red de distribución de agua de mantener un nivel de servicio mínimo bajo condiciones de fallo. Este artículo explora un modelo híbrido que une medidas de resilencia hidráulicas y de teoría de grafos. La propuesta es extender el concepto de distancia geodésica en las tuberías teniendo en cuenta pérdidas de energía asociadas con el caudal. Nuevos algoritmos basados en caminatas aleatorias evaluan rutas hidráulicamente factibles e identifican diferentes niveles de resiliencia en nodos. Aquellos de menor valor en dicha evaluación son analizados en una segunda fase, considerando la disponibilidad y capacidad de sus rutas de abastecimiento.ABSTRACT. Hydraulic resilience can be formulated as a measure of the ability of a water distribution network to maintain a minimum level of service under operational and failure conditions. This paper explores a hybrid approach to bridge the gap between graph-theoretic and hydraulic measures of resilience. We extend the concept of geodesic distance of a pipeline by taking into account energy losses associated with flow. New random-walk algorithms evaluate hydraulically feasible routes and identify nodes with different levels of hydraulic resilience. The nodes with the lowest scores are further analysed by considering the availability and capacity of their supply routes.

A novel hydraulic resilience index for the analysis of water distribution networks (WDN) is presented based on the reserve capacity, a concept extensively studied in transportation network literature. The reserve capacity is defined as a demand multiplier that represents how close a WDN is operating to a minimum service level. A method for calculating the reserve capacity efficiently using Newton's method is presented. Its use is demonstrated with a critical link analysis and a design problem, and compared with another well-established index. The index provides intuitive insight into the behaviour of a WDN that other indexes may not always capture.

An extensive experimental investigation into the pressure management and resilience of three water distribution network configurations is conducted including: fixed topology zones with fixed outlet pressure reducing valves (PRV), fixed topology zones with flow modulating PRVs, and a dynamic topology. Hydraulic data (128S/s) captures the network behaviour under normal conditions and failure, including artificial bursts generated by operating hydrants and a real burst that affected 8,000 properties. Under normal operation, a dynamic topology lowered pressure by 3.1% over a fixed topology with flow modulating PRVs. A dynamic topology maintained the supply of 1,400 properties during the real burst incident.

In this study, we construct a spatio-temporal surface atlas of the developing cerebral cortex, which is an important tool for analysing and understanding normal and abnormal cortical development. In utero Magnetic Resonance Imaging (MRI) of 80 healthy fetuses was performed, with a gestational age range of 21.7 to 38.9 weeks. Topologically correct cortical surface models were extracted from reconstructed 3D MRI volumes. Accurate correspondences were obtained by applying a joint spectral analysis to cortices for sets of subjects close to a specific age. Sulcal alignment was found to be accurate in comparison to spherical demons, a state of the art registration technique for aligning 2D cortical representations (average Fréchet distance ≈ 0.4 mm at 30 weeks). We construct consistent, unbiased average cortical surface templates, for each week of gestation, from age-matched groups of surfaces by applying kernel regression in the spectral domain. These were found to accurately capture the average cortical shape of individuals within the cohort, suggesting a good alignment of cortical geometry. Each spectral embedding and its corresponding cortical surface template provide a dual reference space where cortical geometry is aligned and a vertex-wise morphometric analysis can be undertaken.

The operation of water distribution networks (WDN) with a dynamic topology is a recently pioneered approach for the advanced management of District Metered Areas (DMAs) that integrates novel developments in hydraulic modeling, monitoring, optimization, and control. A common practice for leakage management is the sectorization of WDNs into small zones, called DMAs, by permanently closing isolation valves. This facilitates water companies to identify bursts and estimate leakage levels by measuring the inlet flow for each DMA. However, by permanently closing valves, a number of problems have been created including reduced resilience to failure and suboptimal pressure management. By introducing a dynamic topology to these zones, these disadvantages can be eliminated while still retaining the DMA structure for leakage monitoring. In this paper, a novel optimization method based on sequential convex programming (SCP) is outlined for the control of a dynamic topology with the objective of reducing average zone pressure (AZP). A key attribute for control optimization is reliable convergence. To achieve this, the SCP method we propose guarantees that each optimization step is strictly feasible, resulting in improved convergence properties. By using a null space algorithm for hydraulic analyses, the computations required are also significantly reduced. The optimized control is actuated on a real WDN operated with a dynamic topology. This unique experimental program incorporates a number of technologies set up with the objective of investigating pioneering developments in WDN management. Preliminary results indicate AZP reductions for a dynamic topology of up to 6.5% over optimally controlled fixed topology DMAs.

A dynamic topology aggregates zones in water distribution networks (WDNs) for improved pressure management and resilience to failure. Based on a sequential convex programming (SCP) approach, we propose an optimization method for the control of pressure reducing valves (PRV) in WDNs with dynamic topology. By restricting the SCP iterations to the feasible search space, we show that reliable convergence of the method is achieved. Using an experimental study in a large operational network, the optimization of PRV settings with a dynamic topology is shown to result in pressure reductions of 3.7% compared to optimized PRVs in a closed DMA structure.

A new approach for the operational management of water distribution networks is herein presented, which introduces district metered areas (DMA) with dynamic topology. The approach facilitates the operation of an open and adaptive network that reverts back to the original DMA structure only at night for leakage detection purposes, therefore eliminating the disadvantages of a closed topology such as reduced resilience to failure and suboptimal pressure management. The concept and technology is currently being implemented on a water distribution network in the UK, and a novel optimization method used for its control has been derived that is fast and reliable. (C) 2013 The Authors. Published by Elsevier Ltd. Selection and peer-review under responsibility of the CCWI2013 Committee

This paper presents a novel concept of adaptive water distribution networks with dynamically reconfigurable topology for optimal pressure control, leakage management and improved system resilience. The implementation of District Meter Areas (DMAs) has greatly assisted water utilities in reducing leakage. DMAs segregate water networks into small areas, the flow in and out of each area is monitored and thresholds are derived from the minimum night flow to trigger the leak localization. A major drawback of the DMA approach is the reduced redundancy in network connectivity which has a severe impact on network resilience, incident management and water quality deterioration. The presented approach for adaptively reconfigurable networks integrates the benefits of DMAs for managing leakage with the advantages of large-scale looped networks for increased redundancy in connectivity, reliability and resilience. Self-powered multi-function network controllers are designed and integrated with novel telemetry tools for high-speed time-synchronized monitoring of the dynamic hydraulic conditions. A computationally efficient and robust optimization method based on sequential convex programming is developed and applied for the dynamic topology reconfiguration and pressure control of water distribution networks. An investigation is carried out using an operational network to evaluate the implementation and benefits of the proposed method.

We automatically quantify patterns of normal cortical folding in the developing fetus from in utero MR images (N = 80) over a wide gestational age (GA) range (21.7 to 38.9 weeks). This work on data from healthy subjects represents a first step towards characterising abnormal folding that may be related to pathology, facilitating earlier diagnosis and intervention. The cortical boundary was delineated by automatically segmenting the brain MR image into a number of key structures. This utilised a spatio-temporal atlas as tissue priors in an expectation–maximization approach with second order Markov random field (MRF) regularization to improve the accuracy of the cortical boundary estimate. An implicit high resolution surface was then used to compute cortical folding measures. We validated the automated segmentations with manual delineations and the average surface discrepancy was of the order of 1 mm. Eight curvature-based folding measures were computed for each fetal cortex and used to give summary shape descriptors. These were strongly correlated with GA (R2 = 0.99) confirming the close link between neurological development and cortical convolution. This allowed an age-dependent non-linear model to be accurately fitted to the folding measures. The model supports visual observations that, after a slow initial start, cortical folding increases rapidly between 25 and 30 weeks and subsequently slows near birth. The model allows the accurate prediction of fetal age from an observed folding measure with a smaller error where growth is fastest. We also analysed regional patterns in folding by parcellating each fetal cortex using a nine-region anatomical atlas and found that Gompertz models fitted the change in lobar regions. Regional differences in growth rate were detected, with the parietal and posterior temporal lobes exhibiting the fastest growth, while the cingulate, frontal and medial temporal lobes developed more slowly.