Objectives To assess the role of fetal magnetic resonance imaging (MRI) in detecting associated anomalies in fetuses presenting with mild or moderate isolated ventriculomegaly (VM) undergoing multiplanar ultrasound evaluation of the fetal brain. Methods This was a multicenter, retrospective, cohort study involving 15 referral fetal medicine centers in Italy, the UK and Spain. Inclusion criteria were fetuses affected by isolated mild (ventricular atrial diameter, 10.0–11.9 mm) or moderate (ventricular atrial diameter, 12.0–14.9 mm) VM on ultrasound, defined as VM with normal karyotype and no other additional central nervous system (CNS) or extra‐CNS anomalies on ultrasound, undergoing detailed assessment of the fetal brain using a multiplanar approach as suggested by the International Society of Ultrasound in Obstetrics and Gynecology guidelines for the fetal neurosonogram, followed by fetal MRI. The primary outcome of the study was to report the incidence of additional CNS anomalies detected exclusively on prenatal MRI and missed on ultrasound, while the secondary aim was to estimate the incidence of additional anomalies detected exclusively after birth and missed on prenatal imaging (ultrasound and MRI). Subgroup analysis according to gestational age at MRI (< 24 vs ≥ 24 weeks), laterality of VM (unilateral vs bilateral) and severity of dilatation (mild vs moderate VM) were also performed. Results Five hundred and fifty‐six fetuses with a prenatal diagnosis of isolated mild or moderate VM on ultrasound were included in the analysis. Additional structural anomalies were detected on prenatal MRI and missed on ultrasound in 5.4% (95% CI, 3.8–7.6%) of cases. When considering the type of anomaly, supratentorial intracranial hemorrhage was detected on MRI in 26.7% of fetuses, while polymicrogyria and lissencephaly were detected in 20.0% and 13.3% of cases, respectively. Hypoplasia of the corpus callosum was detected on MRI in 6.7% of cases, while dysgenesis was detected in 3.3%. Fetuses with an associated anomaly detected only on MRI were more likely to have moderate than mild VM (60.0% vs 17.7%; P < 0.001), while there was no significant difference in the proportion of cases with bilateral VM between the two groups (P = 0.2). Logistic regression analysis showed that lower maternal body mass index (adjusted odds ratio (aOR), 0.85 (95% CI, 0.7–0.99); P = 0.030), the presence of moderate VM (aOR, 5.8 (95% CI, 2.6–13.4); P < 0.001) and gestational age at MRI ≥ 24 weeks (aOR, 4.1 (95% CI, 1.1–15.3); P = 0.038) were associated independently with the probability of detecting an associated anomaly on MRI. Associated anomalies were detected exclusively at birth and missed on prenatal imaging in 3.8% of cases. Conclusions The incidence of an associated fetal anomaly missed on ultrasound and detected only on fetal MRI in fetuses with isolated mild or moderate VM undergoing neurosonography is lower than that reported previously. The large majority of these anomalies are difficult to detect on ultrasound. The findings fr...
Applications of optimal transport have recently gained remarkable attention thanks to the computational advantages of entropic regularization. However, in most situations the Sinkhorn approximation of the Wasserstein distance is replaced by a regularized version that is less accurate but easy to differentiate. In this work we characterize the differential properties of the original Sinkhorn distance, proving that it enjoys the same smoothness as its regularized version and we explicitly provide an efficient algorithm to compute its gradient. We show that this result benefits both theory and applications: on one hand, high order smoothness confers statistical guarantees to learning with Wasserstein approximations. On the other hand, the gradient formula allows us to efficiently solve learning and optimization problems in practice. Promising preliminary experiments complement our analysis.
We illustrate some novel contraction and regularizing properties of the Heat flow in metric-measure spaces that emphasize an interplay between Hellinger-Kakutani, Kantorovich-Wasserstein and Hellinger-Kantorvich distances. Contraction properties of Hellinger-Kakutani distances and general Csiszár divergences hold in arbitrary metric-measure spaces and do not require assumptions on the linearity of the flow.When weaker transport distances are involved, we will show that contraction and regularizing effects rely on the dual formulations of the distances and are strictly related to lower Ricci curvature bounds in the setting of RCD(K, ∞) metric measure spaces. Contents2010 Mathematics Subject Classification. Primary: 49Q20, 47D07; Secondary: 30L99.
We present a novel algorithm to estimate the barycenter of arbitrary probability distributions with respect to the Sinkhorn divergence. Based on a Frank-Wolfe optimization strategy, our approach proceeds by populating the support of the barycenter incrementally, without requiring any pre-allocation. We consider discrete as well as continuous distributions, proving convergence rates of the proposed algorithm in both settings. Key elements of our analysis are a new result showing that the Sinkhorn divergence on compact domains has Lipschitz continuous gradient with respect to the Total Variation and a characterization of the sample complexity of Sinkhorn potentials. Experiments validate the effectiveness of our method in practice.
Reinforcement Learning (RL) is emerging as tool for tackling complex control and decision-making problems. However, in high-risk environments such as healthcare, manufacturing, automotive or aerospace, it is often challenging to bridge the gap between an apparently optimal policy learned by an agent and its real-world deployment, due to the uncertainties and risk associated with it. Broadly speaking RL agents face two kinds of uncertainty, 1. aleatoric uncertainty, which reflects randomness or noise in the dynamics of the world, and 2. epistemic uncertainty, which reflects the bounded knowledge of the agent due to model limitations and finite amount of information/data the agent has acquired about the world. These two types of uncertainty carry fundamentally different implications for the evaluation of performance and the level of risk or trust. Yet these aleatoric and epistemic uncertainties are generally confounded as standard and even distributional RL is agnostic to this difference. Here we propose how a distributional approach (UA-DQN) can be recast to render uncertainties by decomposing the net effects of each uncertainty . We demonstrate the operation of this method in grid world examples to build intuition and then show a proof of concept application for an RL agent operating as a clinical decision support system in critical care.
Graph embeddings, wherein the nodes of the graph are represented by points in a continuous space, are used in a broad range of Graph ML applications. The quality of such embeddings crucially depends on whether the geometry of the space matches that of the graph. Euclidean spaces are often a poor choice for many types of real-world graphs, where hierarchical structure and a power-law degree distribution are linked to negative curvature. In this regard, it has recently been shown that hyperbolic spaces and more general manifolds, such as products of constant-curvature spaces and matrix manifolds, are advantageous to approximately match nodes pairwise distances. However, all these classes of manifolds are homogeneous, implying that the curvature distribution is the same at each point, making them unsuited to match the local curvature (and related structural properties) of the graph. In this paper, we study graph embeddings in a broader class of heterogeneous rotationally-symmetric manifolds. By adding a single extra radial dimension to any given existing homogeneous model, we can both account for heterogeneous curvature distributions on graphs and pairwise distances. We evaluate our approach on reconstruction tasks on synthetic and real datasets and show its potential in better preservation of high-order structures and heterogeneous random graphs generation.
We consider the task of sampling from a log-concave probability distribution. This target distribution can be seen as a minimizer of the relative entropy functional defined on the space of probability distributions. The relative entropy can be decomposed as the sum of a functional called the potential energy, assumed to be smooth, and a nonsmooth functional called the entropy. We adopt a Forward Backward (FB) Euler scheme for the discretization of the gradient flow of the relative entropy. This FB algorithm can be seen as a proximal gradient algorithm to minimize the relative entropy over the space of probability measures. Using techniques from convex optimization and optimal transport, we provide a non-asymptotic analysis of the FB algorithm. The convergence rate of the FB algorithm matches the convergence rate of the classical proximal gradient algorithm in Euclidean spaces. The practical implementation of the FB algorithm can be challenging. In practice, the user may choose to discretize the space and work with empirical measures. In this case, we provide a closed form formula for the proximity operator of the entropy.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.