检索结果-内蒙古大学图书馆

arXiv 2020年

作者： Zhang, Molin Xu, Junshen Turk, Esra Abaci Grant, P. Ellen Golland, Polina Adalsteinsson, Elfar Department of Electrical Engineering and Computer Science MIT CambridgeMA United States Fetal-Neonatal Neuroimaging Developmental Science Center Boston Childrens Hospital BostonMA United States Computer Science and Artificial Intelligence Laboratory MIT CambridgeMA United States Harvard Medical School BostonMA United States Institute for Medical Engineering and Science MIT CambridgeMA United States

Fetal MRI is heavily constrained by unpredictable and substantial fetal motion that causes image artifacts and limits the set of viable diagnostic image contrasts. Current mitigation of motion artifacts is predominantly performed by fast, single-shot MRI and retrospective motion correction. Estimation of fetal pose in real time during MRI stands to benefit prospective methods to detect and mitigate fetal motion artifacts where inferred fetal motion is combined with online slice prescription with low-latency decision making. Current developments of deep reinforcement learning (DRL), offer a novel approach for fetal landmarks detection. In this task 15 agents are deployed to detect 15 landmarks simultaneously by DRL. The optimization is challenging, and here we propose an improved DRL that incorporates priors on physical structure of the fetal body. First, we use graph communication layers to improve the communication among agents based on a graph where each node represents a fetal-body landmark. Further, additional reward based on the distance between agents and physical structures such as the fetal limbs is used to fully exploit physical structure. Evaluation of this method on a repository of 3-mm resolution in vivo data demonstrates a mean accuracy of landmark estimation within 10 mm of ground truth as 87.3%, and a mean error of 6.9 mm. The proposed DRL for fetal pose landmark search demonstrates a potential clinical utility for online detection of fetal motion that guides real-time mitigation of motion artifacts as well as health diagnosis during MRI of the pregnant mother. Copyright © 2020, The Authors. All rights reserved.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Time-scale separation based design of biomolecular feedback controllers

Time-scale separation based design of biomolecular feedback ...

引用

IEEE Annual Conference on Decision and Control

作者： Theodore W. Grunberg Domitilla Del Vecchio Department of Electrical Engineering and Computer Science MIT Department of Mechanical Engineering MIT

ISBN: (数字)9781728113982

ISBN: (纸本)9781728113999

Time-scale separation is a powerful property that can be used to simplify control systems design. In this work, we consider the problem of designing biomolecular feedback controllers that provide tracking of slowly varying references and rejection of slowly varying disturbances for nonlinear systems. We propose a design methodology that uses timescale separation to accommodate physical constraints on the implementation of integral control in cellular systems. The main result of this paper gives sufficient conditions under which controllers designed using our time-scale separation methodology have desired asymptotic performance when the reference and disturbance are constant or slowly varying. Our analysis is based on construction of Lyapunov functions for a class of singularly perturbed systems that are dependent on an additional parameter that perturbs the system regularly. When the exogenous inputs are slowly varying, this approach allows us to bound the system trajectories by a function of the regularly perturbing parameter. This bound decays to zero as the parameter's value increases, while an inner-estimate of the region of attraction stays unchanged as this parameter is varied. These results cannot be derived using standard singular perturbation results. We apply our results to an application demonstrating a physically realizable parameter tuning that controls performance.

关键词： Robustness Lyapunov methods Closed loop systems Trajectory computer architecture Perturbation methods Microprocessors

来源：评论

学校读者我要写书评

暂无评论

Semi-supervised learning for fetal brain MRI quality assessment with ROI consistency

arXiv

引用

arXiv 2020年

作者： Xu, Junshen Lala, Sayeri Gagoski, Borjan Turk, Esra Abaci Grant, P. Ellen Golland, Polina Adalsteinsson, Elfar Department of Electrical Engineering and Computer Science MIT CambridgeMA United States Fetal-Neonatal Neuroimaging Developmental Science Center Boston Childrens Hospital BostonMA United States Harvard Medical School BostonMA United States Computer Science and Artificial Intelligence Laboratory MIT CambridgeMA United States Institute for Medical Engineering and Science MIT CambridgeMA United States

Fetal brain MRI is useful for diagnosing brain abnormalities but is challenged by fetal motion. The current protocol for T2-weighted fetal brain MRI is not robust to motion so image volumes are degraded by inter- and intra- slice motion artifacts. Besides, manual annotation for fetal MR image quality assessment are usually time-consuming. Therefore, in this work, a semi-supervised deep learning method that detects slices with artifacts during the brain volume scan is proposed. Our method is based on the mean teacher model, where we not only enforce consistency between student and teacher models on the whole image, but also adopt an ROI consistency loss to guide the network to focus on the brain region. The proposed method is evaluated on a fetal brain MR dataset with 11,223 labeled images and more than 200,000 unlabeled images. Results show that compared with supervised learning, the proposed method can improve model accuracy by about 6% and outperform other state-of-the-art semi-supervised learning methods. The proposed method is also implemented and evaluated on an MR scanner, which demonstrates the feasibility of online image quality assessment and image reacquisition during fetal MR scans. Copyright © 2020, The Authors. All rights reserved.

关键词： Brain

来源：评论

学校读者我要写书评

暂无评论

Accuracy and performance comparison of video action recognition approaches

arXiv

引用

arXiv 2020年

作者： Hutchinson Samsi, Siddharth Arcand, William Bestor, David Bergeron, Bill Byun, Chansup Houle, Micheal Hubbell, Matthew Jones, Micheal Kepner, Jeremy Kirby, Andrew Michaleas, Peter Milechin, Lauren Mullen, Julie Prout, Andrew Rosa, Antonio Reuther, Albert Yee, Charles Gadepally, Vijay MIT Department of Electrical Engineering and Computer Science CambridgeMA United States MIT Lincoln Laboratory Supercomputing Center LexingtonMA United States MIT Department of Earth Atmosphere and Planetary Sciences CambridgeMA United States

Over the past few years, there has been significant interest in video action recognition systems and models. However, direct comparison of accuracy and computational performance results remain clouded by differing training environments, hardware specifications, hyperparameters, pipelines, and inference methods. This article provides a direct comparison between fourteen "off-the-shelf" and state-of-the-art models by ensuring consistency in these training characteristics in order to provide readers with a meaningful comparison across different types of video action recognition algorithms. Accuracy of the models is evaluated using standard Top-1 and Top-5 accuracy metrics in addition to a proposed new accuracy metric. Additionally, we compare computational performance of distributed training from two to sixty-four GPUs on a state-of-the-art HPC system. Copyright © 2020, The Authors. All rights reserved.

关键词： Program processors

来源：评论

学校读者我要写书评

暂无评论

Distinct distances on non-ruled surfaces and between circles

arXiv

引用

arXiv 2020年

作者： Mathialagan, Surya Sheffer, Adam Department of Electrical Engineering and Computer Science Mit MA United States Department of Mathematics Baruch College City University of New York NY United States

We improve the current best bound for distinct distances on non-ruled algebraic surfaces in 3. In particular, we show that n points on such a surface span Ω(n32/39-ϵ) distinct distances, for any ϵ > 0. Our proof adapts the proof of Székely for the planar case, which is based on the crossing lemma. As part of our proof for distinct distances on surfaces, we also obtain new results for distinct distances between circles in 3. Consider two point sets of respective sizes m and n, such that each set lies on a distinct circle in 3. We characterize the cases when the number of distinct distances between the two sets can be O(m + n). This includes a new configuration with a small number of distances. In any other case, we prove that the number of distinct distances is Ω(min {m2/3n2/3, m2, n2}). © 2020, CC BY.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Towards an Efficient and General Framework of Robust Training for Graph Neural Networks

Towards an Efficient and General Framework of Robust Trainin...

引用

IEEE International Conference on Acoustics, Speech and Signal Processing

作者： Kaidi Xu Sijia Liu Pin-Yu Chen Mengshu Sun Caiwen Ding Bhavya Kailkhura Xue Lin Electrical & Computer Engineering Northeastern University Boston USA MIT-IBM Watson AI Lab IBM Research Computer Science and Engineering University of Connecticut Storrs USA Lawrence Livermore National Laboratory Livermore USA

ISBN: (数字)9781509066315

ISBN: (纸本)9781509066322

Graph Neural Networks (GNNs) have made significant advances on several fundamental inference tasks. As a result, there is a surge of interest in using these models for making potentially important decisions in high-regret applications. However, despite GNNs' impressive performance, it has been observed that carefully crafted perturbations on graph structures (or nodes attributes) lead them to make wrong predictions. Presence of these adversarial examples raises serious security concerns. Most of the existing robust GNN design/training methods are only applicable to white-box settings where model parameters are known and gradient based methods can be used by performing convex relaxation of the discrete graph domain. More importantly, these methods are not efficient and scalable which make them infeasible in time sensitive tasks and massive graph datasets. To overcome these limitations, we propose a general framework which leverages the greedy search algorithms and zeroth-order methods to obtain robust GNNs in a generic and an efficient manner. On several applications, we show that the proposed techniques are significantly less computationally expensive and, in some cases, more robust than the state-of-the-art methods making them suitable to large-scale problems which were out of the reach of traditional robust training methods.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Gamma-Ray Burst Light Curve Reconstruction: A Comparative Machine and Deep Learning Analysis

arXiv

引用

arXiv 2024年

作者： Kaushal, A. Manchanda, A. Dainotti, M.G. Deepu, A. Naqi, S. Felix, J. Indoriya, N. Magesh, S.P. Gupta, H. Gupta, K. Madhan, A. Hartmann, D.H. Pollo, A. Bogdan, M. Prochaska, J.X. Fraija, N. Debnath, D. Department of Computer Science and Engineering UIET-H Panjab University Punjab 146001 India Department of Physics Indian Institute of Technology Telangana Hyderabad502284 India National Astronomical Observatory of Japan 2 Chome-21-1 Osawa Mitaka Tokyo181-8588 Japan Department of Astronomical Sciences The Graduate University for Advanced Studies SOKENDAI Shonankokusaimura Miura District Kanagawa Hayama240-0115 Japan Nevada Center for Astrophysics University of Nevada 4505 Maryland Parkway Las VegasNV89154 United States Space Science Institute 4765 Walnut Street Suite B BoulderCO80301 United States Department of Ocean Engineering and Naval Architecture Indian Institute of Technology West Bengal Kharagpur721302 India KNIT Uttar Pradesh Sultanpur228118 India Indian Institute of Technology Kharagpur721302 India Department of Electrical Engineering and Computer Science IISER Madhya Pradesh Bhopal462066 India Indian Institute of Technology Madras600036 India MIT WPU School of Computer Science & Engineering Pune411038 India Department of Computer Science PSIT College Uttar Pradesh Kanpur209305 India Visveswaraya Technological University Karnataka Belagavi590018 India Department of Physics and Astronomy Clemson University ClemsonSC29634 United States Astronomical Observatory of Jagiellonian University in Kraków Orla 171 Kraków30-244 Poland National Centre for Nuclear Research Warsaw02-093 Poland Department of Mathematics University of Wroclaw Wroclaw50-384 Poland University of California Santa Cruz 1156 High Street Santa CruzCA95064 United States National Autonoma University of Mexico Circuito Interior Mexico City04510 Mexico NIT Tripura Agartala799046 India

Gamma-Ray Bursts (GRBs), observed at large redshifts, are probes of the evolution of the Universe and can be used as cosmological tools. To this end, we need tight (with small dispersion) correlations among key parameters. To reduce such a dispersion, we will mitigate gaps in light curves (LCs), including the plateau region, key to building the two-dimensional Dainotti relation between the end time of plateau emission (Ta) to its luminosity (La). We reconstruct LCs using nine models: Multi-Layer Perceptron (MLP), Bi-Mamba, Fourier Transform, Gaussian Process–Random Forest Hybrid (GP-RF), Bidirectional Long Short-Term Memory (Bi-LSTM), Conditional GAN (CGAN), SARIMAX-based Kalman filter, Kolmogorov–Arnold Networks (KANs), and Attention U-Net. These methods are compared to the Willingale model (W07) over a sample of 545 GRBs. MLP and Bi-Mamba outperform other methods, with MLP reducing the plateau parameter uncertainties by 25.9% for log Ta, 28.6% for log Fa, and 37.7% for α (the post-plateau slope in the W07 model), achieving the lowest 5-fold cross validation (CV) mean squared error (MSE) of 0.0275. Bi-Mamba achieved the lowest uncertainty of parameters, a 33.3% reduction in log Ta and log Fa and a 41.9% in α, but with a higher MSE of 0.130. Bi-Mamba brings the lowest outlier percentage for log Ta and log Fa (2.70%), while MLP carry α outliers to 0.900%. The other methods yield MSE values ranging from 0.0339 to 0.174. These improvements in parameter precision are needed to use GRBs as standard candles, investigate theoretical models, and predict GRB redshifts through machine learning. Copyright © 2024, The Authors. All rights reserved.

关键词： Mean square error

来源：评论

学校读者我要写书评

暂无评论

Accelerating the Optimization of Vertical Flow Assay Performance Guided by a Rational Systematic Model-Based Approach

SSRN

引用

SSRN 2022年

作者： Tay, Dousabel M.Y. Kim, Seunghyeon Hao, Yining Yee, Emma H. Jia, Huan Vleck, Sydney M. Chilekwa, Makaya Voldman, Joel Sikes, Hadley D. Department of Chemical Engineering Massachusetts Institute of Technology CambridgeMA02139 United States Antimicrobial Resistance Integrated Research Group Singapore-MIT Alliance for Research and Technology 1 CREATE Way Singapore138602 Singapore Department of Electrical Engineering and Computer Science Massachusetts Institute of Technology CambridgeMA02139 United States Research Laboratory of Electronics Massachusetts Institute of Technology CambridgeMA02139 United States Microsystems Technology Laboratory Massachusetts Institute of Technology CambridgeMA02139 United States

Rapid diagnostic tests (RDTs) have shown to be instrumental in healthcare and disease control. However, they have been plagued by many inefficiencies in the laborious empirical development and optimization process for the attainment of clinically relevant sensitivity. While various studies have sought to model paper-based RDTs, most have relied on continuum-based models that are not necessarily applicable to all operation regimes, and have solely focused on predicting the specific interactions between the antigen and binders. It is also unclear how the model predictions may be utilized for optimizing assay performance. Here, we propose a streamlined and simplified model-based framework, only relying on calibration with a minimal experimental dataset, for the acceleration of assay optimization. We show that our models are capable of recapitulating experimental data across different formats and antigen-binder-matrix combinations. By predicting signals due to both specific and background interactions, our facile approach enables the estimation of several pertinent assay performance metrics such as limit-of-detection, sensitivity, signal-to-noise ratio and difference. We believe that our proposed workflow would be a valuable addition to the toolset of any assay developer, regardless of the amount of resources they have in their arsenal, and aid assay optimization at any stage in their assay development process. © 2022, The Authors. All rights reserved.

关键词： Disease control

来源：评论

学校读者我要写书评

暂无评论

Brain implantation of soft bioelectronics via embryonic development

引用

Nature 2025年 1-11页

作者： Hao Sheng Ren Liu Qiang Li Yichun He Thomas S. Blum Hao Zhao Xin Tang Wenbo Wang Emma Hsiao Paul Le Floch Hao Shen Ariel J. Lee Daniel Solomon Jia Liu Zuwan Lin Lishuai Jin Zheliang Wang Siyi Liu Nanshu Lu Rachael Alice Jonas-Closs James Briggs Xiao Wang Jessica L. Whited John A. Paulson School of Engineering and Applied Sciences Harvard University Boston MA USA Axoft Inc. Cambridge MA USA Electrical Engineering and Computer Science MIT Cambridge MA USA Department of Chemistry and Chemical Biology Harvard University Cambridge MA USA Department of Materials Science and Engineering University of Pennsylvania Philadelphia PA USA Department of Aerospace Engineering and Engineering Mechanics The University of Texas at Austin Austin TX USA Department of Systems Biology Harvard Medical School Boston MA USA Broad Institute of MIT and Harvard Cambridge MA USA Department of Chemistry MIT Cambridge MA USA Department of Stem Cell and Regenerative Biology Harvard University Cambridge MA USA

Developing bioelectronics capable of stably tracking brain-wide, single-cell, millisecond-resolved neural activity in the developing brain is critical for advancing neuroscience and understanding neurodevelopmental disorders. During development, the three-dimensional structure of the vertebrate brain arises from a two-dimensional neural plate1,2. These large morphological changes have previously posed a challenge for implantable bioelectronics to reliably track neural activity throughout brain development3,4,5,6,7,8,9. Here we introduce a tissue-level-soft, submicrometre-thick mesh microelectrode array that integrates into the embryonic neural plate by leveraging the tissue’s natural two-dimensional-to-three-dimensional reconfiguration. As organogenesis progresses, the mesh deforms, stretches and distributes throughout the brain, seamlessly integrating with neural tissue. Immunostaining, gene expression analysis and behavioural testing confirm no adverse effects on brain development or function. This embedded electrode array enables long-term, stable mapping of how single-neuron activity and population dynamics emerge and evolve during brain development. In axolotl models, it not only records neural electrical activity during regeneration but also modulates the process through electrical stimulation.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Synthetic Interventions

arXiv

引用

arXiv 2020年

作者： Agarwal, Anish Shah, Devavrat Shen, Dennis Department of Industrial Engineering & Operations Research Columbia University United States Department of Electrical Engineering & Computer Science MIT United States Department of Data Sciences & Operations USC United States

The synthetic controls (SC) methodology is a prominent tool for policy evaluation in panel data applications. Researchers commonly justify the SC framework with a low-rank matrix factor model that assumes the potential outcomes are described by low-dimensional unit and time specific latent factors. In the recent work of [Aba20], one of the pioneering authors of the SC method posed the question of how the SC framework can be extended to multiple treatments. This article offers one resolution to this open question that we call synthetic interventions (SI). Fundamental to the SI framework is a low-rank tensor factor model, which extends the matrix factor model by including a latent factorization over treatments. Under this model, we propose a generalization of the standard SC-based estimators. We prove the consistency for one instantiation of our approach and provide conditions under which it is asymptotically normal. Moreover, we conduct a representative simulation to study its prediction performance and revisit the canonical SC case study of [ADH10] on the impact of anti-tobacco legislations by exploring related questions not previously investigated. Copyright © 2020, The Authors. All rights reserved.

关键词： Tensors

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：