检索结果-内蒙古大学图书馆

Accurate super-resolution low-field brain MRI

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Iglesias, Juan Eugenio Schleicher, Riana Laguna, Sonia Billot, Benjamin Schaefer, Pamela McKaig, Brenna Goldstein, Joshua N. Sheth, Kevin N. Rosen, Matthew S. Kimberly, W. Taylor Athinoula A. Martinos Center for Biomedical Imaging Massachusetts General Hospital Harvard Medical School BostonMA02114 United States Department of Radiology Massachusetts General Hospital Harvard Medical School BostonMA02114 United States Centre for Medical Image Computing Department of Medical Physics and Biomedical Engineering University College London United Kingdom Computer Science and Artificial Intelligence Laboratory Massachusetts Institute of Technology CambridgeMA02139 United States Department of Neurology and Center for Genomic Medicine Massachusetts General Hospital Harvard Medical School BostonMA02114 United States Zurich Switzerland Department of Emergency Medicine Massachusetts General Hospital Harvard Medical School BostonMA02114 United States Department of Neurology Yale New Haven Hospital New HavenCT06511 United States Department of Physics Harvard University CambridgeMA02138 United States

The recent introduction of portable, low-field MRI (LF-MRI) into the clinical setting has the potential to transform neuroimaging. However, LF-MRI is limited by lower resolution and signal-to-noise ratio, leading to incomplete characterization of brain regions. To address this challenge, recent advances in machine learning facilitate the synthesis of higher resolution images derived from one or multiple lower resolution scans. Here, we report the extension of a machine learning super-resolution (SR) algorithm to synthesize 1 mm isotropic MPRAGE-like scans from LF-MRI T1-weighted and T2-weighted sequences. Our initial results on a paired dataset of LF and high-field (HF, 1.5T-3T) clinical scans show that: (i) application of available automated segmentation tools directly to LF-MRI images falters;but (ii) segmentation tools succeed when applied to SR images with high correlation to gold standard measurements from HF-MRI (e.g., r = 0.85 for hippocampal volume, r = 0.84 for the thalamus, r = 0.92 for the whole cerebrum). This work demonstrates proof-of-principle post-processing image enhancement from lower resolution LF-MRI sequences. These results lay the foundation for future work to enhance the detection of normal and abnormal image findings at LF and ultimately improve the diagnostic performance of LF-MRI. Our tools are publicly available on FreeSurfer (***/). © 2022, CC BY-NC-ND.

关键词： Brain

3D Reconstruction of human brain post-mortem photography and integration of dementia histopathology

学校读者我要写书评

暂无评论

Alzheimer's & Dementia 2023年第S12期19卷

作者： Diana Ortega-Cruz Kimberly S. Bress Harshvardhan Gazula Alberto Rabano Juan Eugenio Iglesias Bryan A Strange CIEN Foundation Queen Sofia Foundation Alzheimer Centre Madrid Spain Centre for Biomedical Technology Universidad Politécnica de Madrid Madrid Spain Vanderbilt University School of Medicine Nashville TN USA Martinos Center for Biomedical Imaging Massachusetts General Hospital and Harvard Medical School Charlestown MA USA Centre for Medical Image Computing Department of Medical Physics and Biomedical Engineering University College London London United Kingdom Computer Science and Artificial Intelligence Laboratory Massachusetts Institute of Technology Cambridge MA USA Centre for Biomedical Technology Universidad Politecnica de Madrid Madrid Spain

Background In Alzheimer’s Disease (AD) and other types of dementia, neuropathological assessment serves as ground truth for diagnosis and validation of novel biomarkers. Such assessment commonly implies inspection of a single histological slide per region of interest. Pathology burden quantification, as well as investigation of its spatial distribution, is thus biased by the variability of tissue sampling in routinary neuropathological practice. Numerous efforts have sought to overcome this lack of generalizability using 3D reconstruction of histology, but available approaches rely on dense sectioning and an MRI scan, which may not be attainable by many brain banks. Method We present a histology mapping pipeline that uses blockface images of brain hemisphere slices for 3D spatial localization. The slices are first geometrically stacked and coarsely aligned using a probabilistic atlas. Next, deep learning contrast synthesis is employed to obtain a high-resolution 3D image of the hemisphere with MRI-like contrast, which is subsequently registered nonlinearly to a standard atlas (MNI). Histological slides are then registered to the corresponding coronal slice within the subject-specific reconstructed hemisphere. Finally, the registrations are concatenated to deform the histology to the standard space. We show the applicability of this pipeline by analyzing the distribution of tau AD pathology along the longitudinal axis of the hippocampus in 26 patients with no significant co-pathologies aside from AD. Result We took advantage of the inter-subject spatial variability observed after 3D slide mapping to achieve extensive sampling along the hippocampal long axis. Our results show an anterior-posterior gradient of tau burden in the hippocampus that is consistent with previous work using ex vivo MRI and dense histological sampling, further supporting the reliability of the pipeline. Conclusion Our pipeline enables studying pathology in 3D using sparse sampling and in absenc

关键词：

Path integral based convolution and pooling for graph neural networks

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Ma, Zheng Xuan, Junyu Wang, Yu Guang Li, Ming Liò, Pietro Department of Physics Princeton University Centre for Artificial Intelligence Faculty of Engineering and Information Technology University of Technology Sydney Australia Max Planck Institute for Mathematics in the Sciences School of Mathematics and Statistics University of New South Wales Department of Educational Technology Zhejiang Normal University Department of Computer Science and Technology University of Cambridge

Graph neural networks (GNNs) extends the functionality of traditional neural networks to graph-structured data. Similar to CNNs, an optimized design of graph convolution and pooling is key to success. Borrowing ideas from physics, we propose a path integral based graph neural networks (PAN) for classification and regression tasks on graphs. Specifically, we consider a convolution operation that involves every path linking the message sender and receiver with learnable weights depending on the path length, which corresponds to the maximal entropy random walk. It generalizes the graph Laplacian to a new transition matrix we call maximal entropy transition (MET) matrix derived from a path integral formalism. Importantly, the diagonal entries of the MET matrix are directly related to the subgraph centrality, thus providing a natural and adaptive pooling mechanism. PAN provides a versatile framework that can be tailored for different graph data with varying sizes and structures. We can view most existing GNN architectures as special cases of PAN. Experimental results show that PAN achieves state-of-the-art performance on various graph classification/regression tasks, including a new benchmark dataset from statistical mechanics we propose to boost applications of GNN in physical sciences. Copyright © 2020, The Authors. All rights reserved.

关键词： Graph neural networks

A Regularization-adaptive Non-negative Latent Factor Analysis-based Model For Recommender Systems

学校读者我要写书评

暂无评论

A Regularization-adaptive Non-negative Latent Factor Analysi...

IEEE International Conference on Human-Machine Systems

作者： Jiufang Chen Xin Luo MengChu Zhou School of Computer Science China West Normal University Nanchong Sichuan China Chinese Academy of Sciences Chongqing Inst. of Green and Intelligent Tech Chongqing China Department of Big Data Analyses Techniques Hengrui (Chongqing) Artificial Intelligence Research Center Cloudwalk Chongqing China Department of Electrical and Computer Engineering New Jersey Inst. of Tech Newark NJ USA

ISBN: (数字)9781728158716

ISBN: (纸本)9781728158723

Non-negative latent factor analysis (NLFA) can high-efficiently extract useful information from high dimensional and sparse (HiDS) matrices often encountered in recommender systems (RSs). However, an NLFA-based model requires careful tuning of regularization coefficients, which is highly expensive in both time and computation. To address this issue, this study proposes an adaptive NLFA-based model whose regularization coefficients become self-adaptive via particle swarm optimization. Experimental results on two HiDS matrices indicate that owing to such self-adaptation, it outperforms an NLFA model in terms of both convergence rate and prediction accuracy for missing data estimation.

关键词： Computational modeling Adaptation models Sparse matrices Analytical models Tuning Particle swarm optimization Big data

Large-scale multi-view subspace clustering in linear time

学校读者我要写书评

暂无评论

arXiv 2019年

作者： Kang, Zhao Zhou, Wangtao Zhao, Zhitong Shao, Junming Han, Meng Xu, Zenglin School of Computer Science and Engineering University of Electronic Science and Technology of China China School of Management and Economics University of Electronic Science and Technology of China China Centre for Artificial Intelligence Peng Cheng Lab Shenzhen518055 China

A plethora of multi-view subspace clustering (MVSC) methods have been proposed over the past few years. Researchers manage to boost clustering accuracy from different points of view. However, many state-of-the-art MVSC algorithms, typically have a quadratic or even cubic complexity, are inefficient and inherently difficult to apply at large scales. In the era of big data, the computational issue becomes critical. To fill this gap, we propose a large-scale MVSC (LMVSC) algorithm with linear order complexity. Inspired by the idea of anchor graph, we first learn a smaller graph for each view. Then, a novel approach is designed to integrate those graphs so that we can implement spectral clustering on a smaller graph. Interestingly, it turns out that our model also applies to single-view scenario. Extensive experiments on various large-scale benchmark data sets validate the effectiveness and efficiency of our approach with respect to state-of-the-art clustering methods. Copyright © 2019, The Authors. All rights reserved.

关键词： Clustering algorithms

Joint segmentation and deformable registration with the presence of the noise using active contour and linear curvature

学校读者我要写书评

暂无评论

AIP Conference Proceedings 2024年第1期3150卷

作者： Nurul Asyiqin Mohd Fauzi Mazlinda Ibrahim Hoo Yann Seong Abdul Kadir Jumaat Lavdie Rada Faculty of Defence Science and Technology National Defence University of Malaysia 57000 Kuala Lumpur Malaysia Centre for Defence Foundation Studies National Defence University of Malaysia 57000 Kuala Lumpur Malaysia School of Mathematical Sciences College of Computing Informatics and Media Universiti Teknologi MARA (UiTM) 40450 Shah Alam Selangor Malaysia Institute for Big Data Analytics and Artificial Intelligence (IBDAAI) Universiti Teknologi MARA (UiTM) 40450 Shah Alam Selangor Malaysia Biomedical Engineering Department Bahcesehir University Besiktas Istanbul Turkey

Medical image processing requires image segmentation and image registration as important diagnostic tools. Image segmentation aims to split structures of interest into distinct components. Meanwhile, image registration finds an ideal transformation between given images. Despite their interrelationship, both tasks are often performed separately. As a result, it faces several problems when undertaken independently. Therefore, the joint models between segmentation and registration have advantages over the separate tasks, particularly when dealing with noisy images. This paper reviewed and compared three existing variational joint segmentation and registration (JSR) models based on active contour without edges and linear curvature. The models are tested on 2D mono-modal and multi-modal images with the presence of white Gaussian noise. The Dice coefficient metric and the registration performance measure are used to evaluate the performance of the models. Numerical results showed that the JSR model using the modified sum of the squared difference outperformed the other two JSR models for noisy mono-modal images. However, the JSR model using the modified normalized gradient fields performed better than the JSR model based on the modified mutual information for noisy multi-modal images.

关键词：

FAMED-Net: A Fast and Accurate Multi-scale End-to-end Dehazing Network

学校读者我要写书评

暂无评论

arXiv 2019年

作者： Zhang, Jing Tao, Dacheng IEEE UBTECH Sydney Artificial Intelligence Centre School of Computer Science in the Faculty of Engineering at the University of Sydney 6 Cleveland St DarlingtonNSW2008 Australia

Single image dehazing is a critical image pre-processing step for subsequent high-level computer vision tasks. However, it remains challenging due to its ill-posed nature. Existing dehazing models tend to suffer from model overcomplexity and computational inefficiency or have limited representation capacity. To tackle these challenges, here we propose a fast and accurate multi-scale end-to-end dehazing network called FAMED-Net, which comprises encoders at three scales and a fusion module to efficiently and directly learn the haze-free image. Each encoder consists of cascaded and densely connected point-wise convolutional layers and pooling layers. Since no larger convolutional kernels are used and features are reused layer-by-layer, FAMED-Net is lightweight and computationally efficient. Thorough empirical studies on public synthetic datasets (including RESIDE) and real-world hazy images demonstrate the superiority of FAMED-Net over other representative state-of-the-art models with respect to model complexity, computational efficiency, restoration accuracy, and cross-set generalization. The code will be made publicly available. Copyright © 2019, The Authors. All rights reserved.

关键词： Image reconstruction

Mitosis domain generalization in histopathology images - The MIDOG challenge

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Aubreville, Marc Stathonikos, Nikolas Bertram, Christof A. Klopfleisch, Robert ter Hoeve, Natalie Ciompi, Francesco Wilm, Frauke Marzahl, Christian Donovan, Taryn A. Maier, Andreas Breen, Jack Ravikumar, Nishant Chung, Youjin Park, Jinah Nateghi, Ramin Pourakpour, Fattaneh Fick, Rutger H.J. Hadj, Saima Ben Jahanifar, Mostafa Rajpoot, Nasir Dexl, Jakob Wittenberg, Thomas Kondo, Satoshi Lafarge, Maxime W. Koelzer, Viktor H. Liang, Jingtang Wang, Yubo Long, Xi Liu, Jingxin Razavi, Salar Khademi, April Yang, Sen Wang, Xiyue Veta, Mitko Breininger, Katharina Technische Hochschule Ingolstadt Ingolstadt Germany Pathology Department UMC Utrecht Netherlands Institute of Pathology University of Veterinary Medicine Vienna Austria Institute of Veterinary Pathology Freie Universität Berlin Berlin Germany Computational Pathology Group Radboud UMC Nijmegen Netherlands Pattern Recognition Lab Friedrich-Alexander-Universität Erlangen-Nürnberg Germany Department of Anatomic Pathology Schwarzman Animal Medical Center New York United States CISTIB Center for Computational Imaging and Simulation Technologies in Biomedicine School of Computing University of Leeds United Kingdom Korea Advanced Institute of Science and Technology Daejeon Korea Republic of Electrical and Electronics Engineering Department Shiraz University of Technology Shiraz Iran Tehran Iran Tribun Health Paris France Tissue Image Analytics Centre Department of Computer Science University of Warwick United Kingdom Fraunhofer-Institute for Integrated Circuits IIS Erlangen Germany Muroran Institute of Technology Hokkaido Japan Department of Pathology and Molecular Pathology University Hospital University of Zurich Zurich Switzerland School of Life Science and Technology Xidian University Shannxi China Histo Pathology Diagnostic Center Shanghai China Xi'an Jiaotong-Liverpool University Suzhou China Electrical Computer and Biomedical Engineering Ryerson University TorontoON Canada Tencent AI Lab Shenzhen518057 China College of Computer Science Sichuan University Chengdu610065 China Medical Image Analysis Group TU Eindhoven Netherlands Department of Artificial Intelligence in Biomedical Engineering Friedrich-Alexander-Universität Erlangen-Nürnberg Germany

The density of mitotic figures within tumor tissue is known to be highly correlated with tumor proliferation and thus is an important marker in tumor grading. Recognition of mitotic figures by pathologists is known to be subject to a strong inter-rater bias, which limits the prognostic value. State-of-the-art deep learning methods can support the expert in this assessment but are known to strongly deteriorate when applied in a different clinical environment than was used for training. One decisive component in the underlying domain shift has been identified as the variability caused by using different whole slide scanners. The goal of the MICCAI MIDOG 2021 challenge has been to propose and evaluate methods that counter this domain shift and derive scanner-agnostic mitosis detection algorithms. The challenge used a training set of 200 cases, split across four scanning systems. As a test set, an additional 100 cases split across four scanning systems, including two previously unseen scanners, were given. The best approaches performed on an expert level, with the winning algorithm yielding an F1 score of 0.748 (CI95: 0.704-0.781). In this paper, we evaluate and compare the approaches that were submitted to the challenge and identify methodological factors contributing to better performance. Copyright © 2022, The Authors. All rights reserved.

关键词： Scanning

Advances in Intelligent Systems and Computing

学校读者我要写书评

暂无评论

Preface

Advances in Intelligent Systems and Computing 2021年第9783642331091期1345 AISC卷 v-vi页

作者： Costa, António Pedro Reis, Luís Paulo Moreira, António Longo, Luca Bryda, Grzegorz Research Centre Ludomedia and University of Aveiro Aveiro Portugal Faculty of Engineering Artificial Intelligence and Computer Science Laboratory [LIACC University of Porto Porto Portugal Research Centre University of Aveiro Aveiro Portugal School of Computer Science Technological University Dublin Dublin Ireland Institute of Sociology Jagiellonian University Kraków Poland