An MRF-UNet Product of Experts for Image Segmentation
School of Biomedical Engineering & Imaging Sciences, KCL（伦敦国王学院生物医学工程与影像科学学院）；Athinoula A. Martinos Center for Biomedical Imaging, MGH and HMS（马萨诸塞总医院与哈佛医学院阿西努拉·A·马蒂诺斯生物医学影像中心）；Wellcome Center for Human Neuroimaging, UCL（伦敦大学学院惠康人类神经影像中心）；Institute of Cognitive Neuroscience, UCL（伦敦大学学院认知神经科学研究所）；Institute of Neurology, UCL（伦敦大学学院神经病学研究所）
Mikael Brudfors, Yaël Balbastre, John Ashburner, Geraint Rees, Parashkev Nachev, Sébastien Ourselin, M. Jorge Cardoso
While convolutional neural networks (CNNs) trained by back-propagation have seen unprecedented success at semantic segmentation tasks, they are known to struggle on out-of-distribution data. Markov random fields (MRFs) on the other hand, encode simpler distributions over labels that, although less flexible than UNets, are less prone to over-fitting. In this paper, we propose to fuse both strategies by computing the product of distributions of a UNet and an MRF. As this product is intractable, we solve for an approximate distribution using an iterative mean-field approach. The resulting MRF-UNet is trained jointly by back-propagation. Compared to other works using conditional random fields (CRFs), the MRF has no dependency on the imaging data, which should allow for less over-fitting. We show on 3D neuroimaging data that this novel network improves generalisation to out-of-distribution samples. Furthermore, it allows the overall number of parameters to be reduced while preserving high accuracy. These results suggest that a classic MRF smoothness prior can allow for less over-fitting when principally integrated into a CNN model. Our implementation is available at https://github.com/balbasty/nitorch.
SAGE: Shape-Adapting Gated Experts for Adaptive Histopathology Image Segmentation
University of Science, VNU-HCM（越南国家大学科学学院）；Trivita AI；University of Technology, VNU-HCM（越南国家大学技术学院）；Michigan State University, USA（美国密歇根州立大学）；The University of Texas at Austin（德克萨斯大学奥斯汀分校）
Deep Ensemble approach for Enhancing Brain Tumor Segmentation in Resource-Limited Settings
University of Ibadan（伊巴丹大学）；African Institute for Mathematical Sciences（非洲数学科学研究所）；University of Cape Town（开普敦大学）；South African National Bioinformatics Institute（南非国家生物信息学研究所）；University of Western Cape（西开普大学）；Carnegie Mellon University（卡内基梅隆大学）；McGill University（麦吉尔大学）；Lawson Health Research Institute（劳森健康研究所）；University of British Columbia（不列颠哥伦比亚大学）
A higher-order MRF based variational model for multiplicative noise reduction
Graz University of Technology（格拉茨工业大学）；University of Science and Technology Beijing（北京科技大学）；Chinese Academy of Sciences（中国科学院）
专家场（Fields of Experts，FoE）图像先验模型是一种基于滤波器的高阶马尔可夫随机场（MRF）模型，在解决图像复原问题中具有良好的效果。受成功的FoE为基础的方法，在这封信中，我们提出了一种新的变分模型的基础上的FoE图像先验模型的乘性噪声降低。由此产生的模型对应于一个非凸最小化问题，这可以解决最近发表的非凸优化算法。基于合成斑点噪声和真实合成孔径雷达（SAR）图像的实验结果表明，我们提出的方法的性能与最好的发表去斑点算法。此外，我们提出的模型还有一个额外的优点，即推理非常有效。{Our基于GPU的实现只需不到1秒即可产生最先进的去斑性能。}
SinGAN-Seg: Synthetic training data generation for medical image segmentation
SimulaMet；Oslo Metropolitan University（奥斯陆城市大学）；Sahlgrenska University Hospital（萨尔格伦斯卡大学医院）；University of Gothenburg（哥德堡大学）；Augere Medical（Augere医疗）；Swedish Medical Group（瑞典医疗集团）
Probing Perfection: The Relentless Art of Meddling for Pulmonary Airway Segmentation from HRCT via a Human-AI Collaboration Based Active Learning Method
Imperial College London（帝国理工学院）；Royal Brompton Hospital（皇家布朗普顿医院）；King's College London（伦敦国王学院）；TECNALIA, Basque Research and Technology Alliance (BRTA)（TECNALIA，巴斯克研究与技术联盟（BRTA））；University of the Basque Country UPV/EHU（巴斯克大学 UPV/EHU）
DBF-UNet: A Two-Stage Framework for Carotid Artery Segmentation with Pseudo-Label Generation
Shenzhen International Graduate School, Tsinghua University（清华大学深圳国际研究生院）；School of Automation, Guangdong University of Technology（广东工业大学自动化学院）；School of Electronic and Computer Engineering, Peking University（北京大学电子与计算机工程学院）
Instance-level quantitative saliency in multiple sclerosis lesion segmentation
Translational Imaging in Neurology (ThINk)（神经学转化成像（ThINk））；Department of Medicine and Biomedical Engineering, University Hospital Basel and University of Basel（医学与生物医学工程系，巴塞尔大学医院和巴塞尔大学）；Department of Neurology, University Hospital Basel（神经病学系，巴塞尔大学医院）；Research Center for Clinical Neuroimmunology and Neuroscience Basel (RC2NB)（巴塞尔临床神经免疫学和神经科学研究中心（RC2NB））；MedGIFT, Institute of Informatics, School of Management, HES-SO Valais-Wallis University of Applied Sciences and Arts Western Switzerland（MedGIFT，信息学院，管理学院，瓦莱-迪-瓦尔省应用科学与艺术西方瑞士大学）；CIBM Center for Biomedical Imaging, Lausanne, Switzerland（洛桑瑞士生物医学成像中心）；Radiology Department, Lausanne University Hospital (CHUV) and University of Lausanne（放射科，洛桑大学医院（CHUV）和洛桑大学）
PULASki: Learning inter-rater variability using statistical distances to improve probabilistic segmentation
Data and Knowledge Engineering Group, Faculty of Computer Science, Otto von Guericke University Magdeburg, Germany（数据与知识工程组，计算机科学系，奥托·冯·格里克大学马格德堡分校）；Genomics Research Centre, Human Technopole, Milan, Italy（基因组研究中心，人类技术pole，米兰，意大利）；Laboratory of Fluid Dynamics and Technical Flows, Otto von Guericke University Magdeburg, Germany（流体动力学与技术流动实验室，奥托·冯·格里克大学马格德堡分校）；Department of Biomedical Magnetic Resonance, Faculty of Natural Sciences, Otto von Guericke University Magdeburg, Germany（生物医学磁共振部门，自然科学系，奥托·冯·格里克大学马格德堡分校）；MedDigit, Department of Neurology, Medical Faculty, University Hospital, Magdeburg, Germany（MedDigit，神经病学部门，医学学院，大学医院，马格德堡，德国）；German Centre for Neurodegenerative Disease, Magdeburg, Germany（德国神经退行性疾病中心，马格德堡，德国）；Centre for Behavioural Brain Sciences, Magdeburg, Germany（行为脑科学中心，马格德堡，德国）；School of Mathematics & Statistics, UNSW Sydney and UNSW Data Science Hub（数学与统计学学院，悉尼大学和UNSW数据科学中心）
本文提出 PULASki，在 Probabilistic UNet 中引入基于统计距离的损失以建模专家标注差异，并在颅内血管和 MS 病灶任务上显著优于四种基线。
Image segmentation of treated and untreated tumor spheroids by Fully Convolutional Networks
HTW Dresden-University of Applied Sciences（德累斯顿应用科学大学）；OncoRay - National Center for Radiation Research in Oncology（OncoRay 肿瘤放射研究国家中心）；Faculty of Medicine and University Hospital Carl Gustav Carus, TUD Dresden University of Technology（德累斯顿工业大学卡尔·古斯塔夫·卡鲁斯医学院及大学医院）；Helmholtz-Zentrum Dresden-Rossendorf（德累斯顿-罗森多夫亥姆霍兹中心）；National Center for Tumor Diseases (NCT)（国家肿瘤疾病中心）；NCT/UCC Dresden（德累斯顿国家肿瘤疾病中心/大学癌症中心）
Weakly Supervised Intracranial Aneurysm Detection and Segmentation in MR angiography via Multi-task UNet with Vesselness Prior
Department of Computer Science and Software Engineering, Concordia University, Montréal, Canada（计算机科学与软件工程系，康科迪亚大学）；NeuroRx Research, Montréal, Canada（NeuroRx研究公司）；Department of Electrical and Computer Engineering, Concordia University, Montréal, Canada（电气与计算机工程系，康科迪亚大学）
Evaluation of state-of-the-art deep learning models in the segmentation of the heart ventricles in parasternal short-axis echocardiograms
Southern Illinois University Edwardsville（南伊利诺伊大学爱德华兹维尔分校）；The Johns Hopkins University School of Medicine（约翰霍普金斯大学医学院）；Columbus Regional Hospital（哥伦布地区医院）；St. Louis Community College（圣路易斯社区学院）
Explainable Convolutional Neural Networks for Retinal Fundus Classification and Cutting-Edge Segmentation Models for Retinal Blood Vessels from Fundus Images
Ahsanullah University of Science and Technology（阿萨努拉科技大学）
Cross-modality (CT-MRI) prior augmented deep learning for robust lung tumor segmentation from small MR datasets
Memorial Sloan-Kettering Cancer Center（纪念斯隆-凯特琳癌症中心）
缺乏大型专家注释的MR数据集使得训练深度学习模型变得困难。因此，开发了一种跨模态（MR-CT）深度学习分割方法，该方法使用通过变换专家分割的CT图像产生的伪MR图像来增强训练数据。分析了28例非小细胞肺癌患者的81例MRI T2加权像。将CT转换为类似于T2 w MRI的伪MR图像的跨模态先验编码作为生成对抗深度学习模型进行学习。该模型增强了来自6个专家分割的T2 w MR患者扫描的训练数据，其中377个伪MRI来自非小细胞肺癌CT患者扫描，从癌症成像档案中获得。采用批量归一化实现的二维Unet被训练以分割来自T2 w MRI的肿瘤。该方法以（a）标准数据增强和两种最先进的跨模态伪MR增强以及（b）两种分割网络为基准。分割精度计算使用骰子相似系数（DSC），Hausdroff距离度量，和体积比。所提出的方法在伪MR图像和T2 w MR图像之间的强度分布中产生了最低的统计变异性，测量为0.069的Kullback-Leibler发散。该方法产生了最高的分割精度，DSC为0.75，测试数据集上的Hausdroff距离最低。这种方法产生了与专家高度相似的肿瘤生长估计（P = 0.37）。开发了一种新的深度学习MR分割，其通过利用学习的跨模态先验来增强训练，克服了从小数据集学习鲁棒模型的限制。结果表明，该方法的可行性和相应的改进国家的最先进的方法。
DeepEdit: Deep Editable Learning for Interactive Segmentation of 3D Medical Images
School of Biomedical Engineering & Imaging Sciences, King’s College London（伦敦国王学院生物医学工程与影像科学学院）；NVIDIA（英伟达）；Semmelweis University（塞麦尔维斯大学）；EBATINCA, S.L.（EBATINCA有限公司）；Beth Israel Deaconess Medical Center（贝斯以色列女执事医疗中心）；Isomics, Inc.（Isomics公司）
Automated Detection of Multiple Sclerosis Lesions on 7-tesla MRI Using U-net and Transformer-based Segmentation
Department of Neurology, University of Maryland School of Medicine（马里兰大学医学院神经内科）；Baltimore VA Medical Center, VA Maryland Healthcare System（巴尔的摩退伍军人事务医疗中心，马里兰州退伍军人医疗系统）
