The brain white matter consists of a set of tracts that connect distinct regions of the brain. Segmentation of these tracts is often needed for clinical and research studies. Diffusion-weighted MRI offers unique contrast to delineate these tracts. However, existing segmentation methods rely on intermediate computations such as tractography or estimation of fiber orientation density. These intermediate computations, in turn, entail complex computations that can result in unnecessary errors. Moreover, these intermediate computations often require dense multi-shell measurements that are unavailable in many clinical and research applications. As a result, current methods suffer from low accuracy and poor generalizability. Here, we propose a new deep learning method that segments these tracts directly from the diffusion MRI data, thereby sidestepping the intermediate computation errors. Our experiments show that this method can achieve segmentation accuracy that is on par with the state of the art methods (mean Dice Similarity Coefficient of 0.826). Compared with the state of the art, our method offers far superior generalizability to undersampled data that are typical of clinical studies and to data obtained with different acquisition protocols. Moreover, we propose a new method for detecting inaccurate segmentations and show that it is more accurate than standard methods that are based on estimation uncertainty quantification. The new methods can serve many critically important clinical and scientific applications that require accurate and reliable non-invasive segmentation of white matter tracts.
相關內容
Multimodality of the likelihood in Gaussian mixtures is a well-known problem. The choice of the initial parameter vector for the numerical optimizer may affect whether the optimizer finds the global maximum, or gets trapped in a local maximum of the likelihood. We propose to use Hamiltonian Monte Carlo (HMC) to explore the part of the parameter space which has a high likelihood. Each sampled parameter vector is used as the initial value for quasi-Newton optimizer, and the resulting sample of (maximum) likelihood values is used to determine if the likelihood is multimodal. We use a single simulated data set from a three component bivariate mixture to develop and test the method. We use state-of-the-art HCM software, but experience difficulties when trying to directly apply HMC to the full model with 15 parameters. To improve the mixing of the Markov Chain we explore various tricks, and conclude that for the dataset at hand we have found the global maximum likelihood estimate.
Interactions with large language models have led to the suggestion that these models may soon be conscious. From the perspective of neuroscience, this position is difficult to defend. For one, the inputs to large language models lack the embodied, embedded information content characteristic of our sensory contact with the world around us. Secondly, the architecture of large language models is missing key features of the thalamocortical system that have been linked to conscious awareness in mammals. Finally, the evolutionary and developmental trajectories that led to the emergence of living conscious organisms arguably have no parallels in artificial systems as envisioned today. The existence of living organisms depends on their actions, and their survival is intricately linked to multi-level cellular, inter-cellular, and organismal processes culminating in agency and consciousness.
The occurrence of extreme events like heavy precipitation or storms at a certain location often shows a clustering behaviour and is thus not described well by a Poisson process. We construct a general model for the inter-exceedance times in between such events which combines different candidate models for such behaviour. This allows us to distinguish data generating mechanisms leading to clusters of dependent events with exponential inter-exceedance times in between clusters from independent events with heavy-tailed inter-exceedance times, and even allows us to combine these two mechanisms for better descriptions of such occurrences. We investigate a modification of the Cram\'er-von Mises distance for the purpose of model fitting. An application to mid-latitude winter cyclones illustrates the usefulness of our work.
Recent wildfires in Australia have led to considerable economic loss and property destruction, and there is increasing concern that climate change may exacerbate their intensity, duration, and frequency. hazard quantification for extreme wildfires is an important component of wildfire management, as it facilitates efficient resource distribution, adverse effect mitigation, and recovery efforts. However, although extreme wildfires are typically the most impactful, both small and moderate fires can still be devastating to local communities and ecosystems. Therefore, it is imperative to develop robust statistical methods to reliably model the full distribution of wildfire spread. We do so for a novel dataset of Australian wildfires from 1999 to 2019, and analyse monthly spread over areas approximately corresponding to Statistical Areas Level 1 and 2 (SA1/SA2) regions. Given the complex nature of wildfire ignition and spread, we exploit recent advances in statistical deep learning and extreme value theory to construct a parametric regression model using graph convolutional neural networks and the extended generalized Pareto distribution, which allows us to model wildfire spread observed on an irregular spatial domain. We highlight the efficacy of our newly proposed model and perform a wildfire hazard assessment for Australia and population-dense communities, namely Tasmania, Sydney, Melbourne, and Perth.
An important challenge in Machine Learning compilers like XLA is multi-pass optimization and analysis. There has been recent interest chiefly in XLA target-dependent optimization on the graph-level, subgraph-level, and kernel-level phases. We specifically focus on target-independent optimization XLA HLO pass ordering: our approach aims at finding the optimal sequence of compiler optimization passes, which is decoupled from target-dependent optimization. However, there is little domain specific study in pass ordering for XLA HLO. To this end, we propose introducing deep Reinforcement Learning (RL) based search for optimal XLA HLO pass ordering. We also propose enhancements to the deep RL algorithms to further improve optimal search performance and open the research direction for domain-specific guidance for RL. We create an XLA Gym experimentation framework as a tool to enable RL algorithms to interact with the compiler for passing optimizations and thereby train agents. Overall, in our experimentation we observe an average of $13.3\%$ improvement in operation count reduction on a benchmark of GPT-2 training graphs and $10.4\%$ improvement on a diverse benchmark including GPT-2, BERT, and ResNet graphs using the proposed approach over the compiler's default phase ordering.
The ensemble Gaussian mixture filter (EnGMF) is a powerful filter for highly non-Gaussian and non-linear models that has practical utility in the case of a small number of samples, and theoretical convergence to full Bayesian inference in the ensemble limit. We aim to increase the utility of the EnGMF by introducing an ensemble-local notion of covariance into the kernel density estimation (KDE) step for the prior distribution. We prove that in the Gaussian case, our new ensemble-localized KDE technique is exactly the same as more traditional KDE techniques. We also show an example of a non-Gaussian distribution that can fail to be approximated by canonical KDE methods, but can be approximated exactly by our new KDE technique. We showcase our new KDE technique on a simple bivariate problem, showing that it has nice qualitative and quantitative properties, and significantly improves the estimate of the prior and posterior distributions for all ensemble sizes tested. We additionally show the utility of the proposed methodology for sequential filtering for the Lorenz '63 equations, achieving a significant reduction in error, and less conservative behavior in the uncertainty estimate with respect to traditional techniques.
The pursuit of scientific knowledge strongly depends on the ability to reproduce and validate research results. It is a well-known fact that the scientific community faces challenges related to transparency, reliability, and the reproducibility of empirical published results. Consequently, the design and preparation of reproducible artifacts has a fundamental role in the development of science. Reproducible artifacts comprise comprehensive documentation, data, and code that enable replication and validation of research findings by others. In this work, we discuss a methodology to construct reproducible artifacts based on Docker. Our presentation centers around the preparation of an artifact to be submitted to scientific venues that encourage or require this process. This report's primary audience are scientists working with empirical computer science; however, we believe that the presented methodology can be extended to other technology-oriented empirical disciplines.
This book chapter delves into the pressing need to "queer" the ethics of AI to challenge and re-evaluate the normative suppositions and values that underlie AI systems. The chapter emphasizes the ethical concerns surrounding the potential for AI to perpetuate discrimination, including binarism, and amplify existing inequalities due to the lack of representative datasets and the affordances and constraints depending on technology readiness. The chapter argues that a critical examination of the neoliberal conception of equality that often underpins non-discrimination law is necessary and cannot stress more the need to create alternative interdisciplinary approaches that consider the complex and intersecting factors that shape individuals' experiences of discrimination. By exploring such approaches centering on intersectionality and vulnerability-informed design, the chapter contends that designers and developers can create more ethical AI systems that are inclusive, equitable, and responsive to the needs and experiences of all individuals and communities, particularly those who are most vulnerable to discrimination and harm.
Moire patterns occur when capturing images or videos on screens, severely degrading the quality of the captured images or videos. Despite the recent progresses, existing video demoireing methods neglect the physical characteristics and formation process of moire patterns, significantly limiting the effectiveness of video recovery. This paper presents a unified framework, DTNet, a direction-aware and temporal-guided bilateral learning network for video demoireing. DTNet effectively incorporates the process of moire pattern removal, alignment, color correction, and detail refinement. Our proposed DTNet comprises two primary stages: Frame-level Direction-aware Demoireing and Alignment (FDDA) and Tone and Detail Refinement (TDR). In FDDA, we employ multiple directional DCT modes to perform the moire pattern removal process in the frequency domain, effectively detecting the prominent moire edges. Then, the coarse and fine-grained alignment is applied on the demoired features for facilitating the utilization of neighboring information. In TDR, we propose a temporal-guided bilateral learning pipeline to mitigate the degradation of color and details caused by the moire patterns while preserving the restored frequency information in FDDA. Guided by the aligned temporal features from FDDA, the affine transformations for the recovery of the ultimate clean frames are learned in TDR. Extensive experiments demonstrate that our video demoireing method outperforms state-of-the-art approaches by 2.3 dB in PSNR, and also delivers a superior visual experience.
Although measuring held-out accuracy has been the primary approach to evaluate generalization, it often overestimates the performance of NLP models, while alternative approaches for evaluating models either focus on individual tasks or on specific behaviors. Inspired by principles of behavioral testing in software engineering, we introduce CheckList, a task-agnostic methodology for testing NLP models. CheckList includes a matrix of general linguistic capabilities and test types that facilitate comprehensive test ideation, as well as a software tool to generate a large and diverse number of test cases quickly. We illustrate the utility of CheckList with tests for three tasks, identifying critical failures in both commercial and state-of-art models. In a user study, a team responsible for a commercial sentiment analysis model found new and actionable bugs in an extensively tested model. In another user study, NLP practitioners with CheckList created twice as many tests, and found almost three times as many bugs as users without it.
小貼士
登錄享
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191