检索结果-内蒙古大学图书馆

In vivo high-resolution multimodal nonlinear optical microscopy of spinal cord in mice

学校读者我要写书评

暂无评论

In vivo high-resolution multimodal nonlinear optical microsc...

Advanced Chemical Microscopy for Life Science and Translational Medicine 2020

作者： Wu, Wanjie Qin, Zhongya Wu, Junqiang Chen, Congping Chen, Weitao Liu, Kai Qu, Jianan Y. Biophotonics Research Laboratory Department of Electronic and Computer Engineering Hong Kong University of Science and Technology Clear Water Bay Kowloon Hong Kong Division of Life Science State Key Laboratory of Molecular Neuroscience Hong Kong University of Science and Technology Clear Water Bay Kowloon Hong Kong Center of Systems Biology and Human Health Hong Kong University of Science and Technology Clear Water Bay Kowloon Hong Kong

ISBN: (纸本)9781510632677

Chronic in vivo optical imaging of the spinal cord is an effective way to study the biological processes during and after spinal cord injury (SCI) in mouse models. It normally relies on an implanted spinal chamber to provide continuous optical access to the spinal cord. However, the chronic window consists of multiple layers of transparent materials with various optical properties and irregular thickness, which induce large optical aberration. Therefore, the image quality of multiphoton microscopy as well as the precision of femtosecond laser axotomy were dramatically degraded. In this work, we developed an adaptive optics (AO) microscope system integrating stimulated Raman scattering (SRS) and twophoton excited fluorescence (TPEF). Using our system, the aberrations induced by the spinal cord window were measured and compensated accordingly, enabling both high-resolution imaging and precise laser axotomy of the mouse spinal cord. © COPYRIGHT SPIE. Downloading of the abstract is permitted for personal use only.

关键词： Adaptive optics

Towards universal end-to-end affect recognition from multilingual speech by convnets

学校读者我要写书评

暂无评论

arXiv 2019年

作者： Bertero, Dario Kampman, Onno Fung, Pascale Human Language Technology Center Department of Electronic and Computer Engineering Hong Kong University of Science and Technology Clear Water Bay Hong Kong

We propose an end-to-end affect recognition approach using a Convolutional Neural Network (CNN) that handles multiple languages, with applications to emotion and personality recognition from speech. We lay the foundation of a universal model that is trained on multiple languages at once. As affect is shared across all languages, we are able to leverage shared information between languages and improve the overall performance for each one. We obtained an average improvement of 12.8% on emotion and 10.1% on personality when compared with the same model trained on each language only. It is endto- end because we directly take narrow-band raw waveforms as input. This allows us to accept as input audio recorded from any source and to avoid the overhead and information loss of feature extraction. It outperforms a similar CNN using spectrograms as input by 12.8% for emotion and 6.3% for personality, based on Fscores. Analysis of the network parameters and layers activation shows that the network learns and extracts significant features in the first layer, in particular pitch, energy and contour variations. Subsequent convolutional layers instead capture language-specific representations through the analysis of supra-segmental features. Our model represents an important step for the development of a fully universal affect recognizer, able to recognize additional descriptors, such as stress, and for the future implementation into affective interactive systems. Copyright © 2019, The Authors. All rights reserved.

关键词： Convolution

Evaluation of OpenAI o1: Opportunities and Challenges of AGI

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Zhong, Tianyang Liu, Zhengliang Pan, Yi Zhang, Yutong Zhou, Yifan Liang, Shizhe Wu, Zihao Lyu, Yanjun Shu, Peng Yu, Xiaowei Cao, Chao Jiang, Hanqi Chen, Hanxu Li, Yiwei Chen, Junhao Hu, Huawen Liu, Yihen Zhao, Huaqin Xu, Shaochen Dai, Haixing Zhao, Lin Zhang, Ruidong Zhao, Wei Yang, Zhenyuan Chen, Jingyuan Wang, Peilong Ruan, Wei Wang, Hui Zhao, Huan Zhang, Jing Ren, Yiming Qin, Shihuan Chen, Tong Li, Jiaxi Zidan, Arif Hassan Jahin, Afrar Chen, Minheng Xia, Sichen Holmes, Jason Zhuang, Yan Wang, Jiaqi Xu, Bochen Xia, Weiran Yu, Jichao Tang, Kaibo Yang, Yaxuan Sun, Bolun Yang, Tao Lu, Guoyu Wang, Xianqiao Chai, Lilong Li, He Lu, Jin Sun, Lichao Zhang, Xin Ge, Bao Hu, Xintao Zhang, Lian Zhou, Hua Zhang, Lu Zhang, Shu Liu, Ninghao Jiang, Bei Kong, Linglong Xiang, Zhen Ren, Yudan Liu, Jun Jiang, Xi Bao, Yu Zhang, Wei Li, Xiang Li, Gang Liu, Wei Shen, Dinggang Sikora, Andrea Zhai, Xiaoming Zhu, Dajiang Zhang, Tuo Liu, Tianming Department of Mathematical and Statistical Sciences University of Alberta Edmonton Canada School of Computing University of Georgia GA United States Institute of Medical Research Northwestern Polytechnical University Xi'An China College of Arts and Sciences University of Georgia Athens United States Institute of Plant Breeding Genetics & Genomics University of Georgia AthensGA United States Department of Computer Science and Engineering University of Texas ArlingtonTX United States The Lamar Dodd School of Art University of Georgia GA United States School of Computer Science Northwestern Polytechnical University Xi'An China School of Automation Northwestern Polytechnical University Xi'An China University of California Los AngelesCA United States Department of Radiology The Second Xiangya Hospital Central South University Changsha China Clinical Research Center for Medical Imaging in Hunan Province Changsha China Institute of Biomedical and Health Engineering Shenzhen Institutes of Advanced Technology Chinese Academy of Sciences Shenzhen China Guanghua School of Management Peking University Beijing China Department of Radiation Oncology Mayo Clinic PhoenixAZ United States Second Language Acquisition and Teaching University of Arizona TucsonAZ United States School of Life Science and Technology University of Electronic Science and Technology of China Chengdu China The First Hospital of Hebei Medical University Hebei Shijiazhuang China School of Computer and Cyber Sciences Augusta University AugustaGA United States School of Physics and Information Technology Shaanxi Normal University Xi'An China School of Future Technology South China University of Technology Guangzhou China Department of Radiology and BRIC University of North Carolina Chapel HillNC United States Department of Educational Psychology University of Georgia AthensGA United States Johns Hopkins University BaltimoreMD United States School of Architecture Tsinghua Universit

This comprehensive study evaluates the performance of OpenAI's o1-preview large language model across a diverse array of complex reasoning tasks, spanning multiple domains, including computer science, mathematics, natural sciences, medicine, linguistics, and social sciences. Through rigorous testing, o1-preview demonstrated remarkable capabilities, often achieving human-level or superior performance in areas ranging from coding challenges to scientific reasoning and from language processing to creative problem-solving. Key findings include: • 83.3% success rate in solving complex competitive programming problems, surpassing many human experts. • Superior ability in generating coherent and accurate radiology reports, outperforming other evaluated models. • 100% accuracy in high school-level mathematical reasoning tasks, providing detailed step-by-step solutions. • Advanced natural language inference capabilities across general and specialized domains like medicine. • Impressive performance in chip design tasks, outperforming specialized models in areas such as EDA script generation and bug analysis. • Remarkable proficiency in anthropology and geology, demonstrating deep understanding and reasoning in these specialized fields. • Strong capabilities in quantitative investing. O1 has comprehensive financial knowledge and statistical modeling skills. • Effective performance in social media analysis, including sentiment analysis and emotion recognition. The model excelled particularly in tasks requiring intricate reasoning and knowledge integration across various fields. While some limitations were observed, including occasional errors on simpler problems and challenges with certain highly specialized concepts, the overall results indicate significant progress towards artificial general intelligence. This evaluation not only highlights o1-preview's current strengths and limitations but also identifies crucial areas for future development, including multi-modal integratio

关键词： Emotion Recognition

Assessing Trustworthy AI in Times of COVID-19: Deep Learning for Predicting a Multiregional Score Conveying the Degree of Lung Compromise in COVID-19 Patients

学校读者我要写书评

暂无评论

IEEE Transactions on Technology and Society

IEEE Transactions on technology and Society 2022年第4期3卷 272-289页

作者： Allahabadi, Himanshi Amann, Julia Balot, Isabelle Beretta, Andrea Binkley, Charles Bozenhard, Jonas Bruneault, Frederick Brusseau, James Candemir, Sema Cappellini, Luca Alessandro Chakraborty, Subrata Cherciu, Nicoleta Cociancig, Christina Coffee, Megan Ek, Irene Espinosa-Leal, Leonardo Farina, Davide Fieux-Castagnet, Genevieve Frauenfelder, Thomas Gallucci, Alessio Giuliani, Guya Golda, Adam Van Halem, Irmhild Hildt, Elisabeth Holm, Sune Kararigas, Georgios Krier, Sebastien A. Kuhne, Ulrich Lizzi, Francesca Madai, Vince I. Markus, Aniek F. Masis, Serg Mathez, Emilie Wiinblad Mureddu, Francesco Neri, Emanuele Osika, Walter Ozols, Matiss Panigutti, Cecilia Parent, Brendan Pratesi, Francesca Moreno-Sanchez, Pedro A. Sartor, Giovanni Savardi, Mattia Signoroni, Alberto Sormunen, Hanna-Maria Spezzatti, Andy Srivastava, Adarsh Stephansen, Annette F. Theng, Lau Bee Tithi, Jesmin Jahan Tuominen, Jarno Umbrello, Steven Vaccher, Filippo Vetter, Dennis Westerlund, Magnus Wurth, Renee Zicari, Roberto V. Ey Netherlands Enterprise Intelligence Department Amsterdam1083 HP Netherlands Eth Zurich Health Ethics and Policy Lab Department of Health Sciences and Technology Zürich8092 Switzerland Center for Diplomatic and Strategic Studies Postgraduate Studies in Diplomacy and International Relations Paris75015 France Pisa56124 Italy Hackensack Meridian Health Bioethics Center EdisonNJ08820 United States University of Oxford Faculty of Philosophy OxfordOX2 6GG United Kingdom Collège André- Laurendeau Philosophie Department MontrealQCH8N 2J4 Canada Université du Québec À Montréal École des Médias MontrealQCH2L 2C4 Canada Pace University Philosophy Department New YorkNY10038 United States The Ohio State University Wexner Medical Center Department of Radiology ColumbusOH43210 United States Humanitas Research Hospital Department of Radiology Milan20089 Italy Humanitas University Department of Biomedical Sciences Milan20089 Italy University of New England Faculty of Science Agriculture Business and Law ArmidaleNSW2351 Australia University of Technology Sydney Faculty of Engineering and Information Technology SydneyNSW2007 Australia Scuola Superiore Sant'Anna European Centre of Excellence on the Regulation of Robotics and Ai Pisa56127 Italy University of Bremen Group of Computer Architecture Bremen28359 Germany New York University Grossman School of Medicine Division of Infectious Diseases and Immunology Department of Medicine New YorkNY10016 United States Digital Institute Ai Research Section Stockholm16731 Sweden Arcada University of Applied Sciences Department of Business Management and Analytics Helsinki00550 Finland University of Brescia Radiological Sciences and Public Health Department of Medical and Surgical Specialties Brescia25121 Italy Sncf Reseau Sa Ethique Groupe La Plaine93418 France Institute of Diagnostic and Interventional Radiology University Hospital Zurich Zürich8091 Switzerland Eindhoven University of Tech

This article's main contributions are twofold: 1) to demonstrate how to apply the general European Union's High-Level Expert Group's (EU HLEG) guidelines for trustworthy AI in practice for the domain of healthcare and 2) to investigate the research question of what does 'trustworthy AI' mean at the time of the COVID-19 pandemic. To this end, we present the results of a post-hoc self-assessment to evaluate the trustworthiness of an AI system for predicting a multiregional score conveying the degree of lung compromise in COVID-19 patients, developed and verified by an interdisciplinary team with members from academia, public hospitals, and industry in time of pandemic. The AI system aims to help radiologists to estimate and communicate the severity of damage in a patient's lung from Chest X-rays. It has been experimentally deployed in the radiology department of the ASST Spedali Civili clinic in Brescia, Italy, since December 2020 during pandemic time. The methodology we have applied for our post-hoc assessment, called Z-Inspection®, uses sociotechnical scenarios to identify ethical, technical, and domain-specific issues in the use of the AI system in the context of the pandemic. © 2020 IEEE.

关键词： COVID-19

Deep Separable Spatiotemporal Learning for Fast Dynamic Cardiac MRI

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Wang, Zi Xiao, Min Zhou, Yirong Wang, Chengyan Wu, Naiming Li, Yi Gong, Yiwen Chang, Shufu Chen, Yinyin Zhu, Liuhong Zhou, Jianjun Cai, Congbo Wang, He Guo, Di Yang, Guang Qu, Xiaobo Department of Electronic Science Intelligent Medical Imaging R&D Center Fujian Provincial Key Laboratory of Plasma and Magnetic Resonance National Institute for Data Science in Health and Medicine Xiamen University China Institute of Artificial Intelligence Xiamen University China Human Phenome Institute Fudan University China Department of Imaging Xiamen Cardiovascular Hospital of Xiamen University School of Medicine Xiamen University China Department of Cardiovascular Medicine Heart Failure Center Ruijin Hospital Lu Wan Branch Shanghai Jiaotong University School of Medicine China Shanghai Institute of Cardiovascular Diseases Zhongshan Hospital Fudan University China Department of Radiology Zhongshan Hospital Fudan University Department of Medical Imaging Shanghai Medical School Shanghai Institute of Medical Imaging China Fujian Province Key Clinical Specialty Construction Project Medical Imaging Department Xiamen Key Laboratory of Clinical Transformation of Imaging Big Data and Artificial Intelligence China School of Computer and Information Engineering Xiamen University of Technology China Department of Bioengineering and Imperial-X Imperial College London United Kingdom Department of Bioengineering Imperial College London United Kingdom

Dynamic magnetic resonance imaging (MRI) plays an indispensable role in cardiac diagnosis. To enable fast imaging, the k-space data can be undersampled but the image reconstruction poses a great challenge of high-dimensional processing. This challenge necessitates extensive training data in deep learning reconstruction methods. In this work, we propose a novel and efficient approach, leveraging a dimension-reduced separable learning scheme that can perform exceptionally well even with highly limited training data. We design this new approach by incorporating spatiotemporal priors into the development of a Deep Separable Spatiotemporal Learning network (DeepSSL), which unrolls an iteration process of a 2D spatiotemporal reconstruction model with both temporal low-rankness and spatial sparsity. Intermediate outputs can also be visualized to provide insights into the network behavior and enhance interpretability. Extensive results on cardiac cine datasets demonstrate that the proposed DeepSSL surpasses state-of-the-art methods both visually and quantitatively, while reducing the demand for training cases by up to 75%. Additionally, its preliminary adaptability to unseen cardiac patients has been verified through a blind reader study conducted by experienced radiologists and cardiologists. Furthermore, DeepSSL enhances the accuracy of the downstream task of cardiac segmentation and exhibits robustness in prospectively undersampled real-time cardiac MRI. Copyright © 2024, The Authors. All rights reserved.

关键词： Image reconstruction

CMR×Recon: An open cardiac MRI dataset for the competition of accelerated image reconstruction

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Wang, Chengyan Lyu, Jun Wang, Shuo Qin, Chen Guo, Kunyuan Zhang, Xinyu Yu, Xiaotong Li, Yan Wang, Fanwen Jin, Jianhua Shi, Zhang Xu, Ziqiang Tian, Yapeng Hua, Sha Chen, Zhensen Liu, Meng Sun, Mengting Kuang, Xutong Wang, Kang Wang, Haoran Li, Hao Chu, Yinghua Yang, Guang Bai, Wenjia Zhuang, Xiahai Wang, He Qin, Jing Qu, Xiaobo Human Phenome Institute Fudan University Shanghai China School of Nursing The Hong Kong Polytechnic University Hong Kong Digital Medical Research Center School of Basic Medical Sciences Fudan University Shanghai China Department of Electrical and Electronic Engineering & I-X Imperial College London United Kingdom Department of Electronic Science Fujian Provincial Key Laboratory of Plasma and Magnetic Resonance National Institute for Data Science in Health and Medicine Institute of Artificial Intelligence Xiamen University Xiamen China Department of Radiology Ruijin Hospital Shanghai Jiao Tong University School of Medicine Shanghai China Department of Bioengineering/˜Imperial-X Imperial College London United Kingdom School of Data Science Fudan University Shanghai China Department of Radiology Zhongshan Hospital Fudan University Shanghai China School of Health Science and Engineering University of Shanghai for Science and Technology Shanghai China Department of Computer Science The University of Texas Dallas United States Department of Cardiovascular Medicine Ruijin Hospital Lu Wan Branch Shanghai Jiao Tong University School of Medicine Shanghai China Institute of Science and Technology for Brain-Inspired Intelligence Fudan University Shanghai200433 China Simens Healthineers Ltd. China Department of Brain Sciences Imperial College London London United Kingdom Department of Computing Imperial College London London United Kingdom

Cardiac magnetic resonance imaging (CMR) has emerged as a valuable diagnostic tool for cardiac diseases. However, a limitation of CMR is its slow imaging speed, which causes patient discomfort and introduces artifacts in the images. There has been growing interest in deep learning-based CMR imaging algorithms that can reconstruct high-quality images from highly under-sampled k-space data. However, the development of deep learning methods requires large training datasets, which have not been publicly available for CMR. To address this gap, we released a dataset that includes multi-contrast, multi-view, multi-slice and multi-coil CMR imaging data from 300 subjects. Imaging studies include cardiac cine and mapping sequences. Manual segmentations of the myocardium and chambers of all the subjects are also provided within the dataset. Scripts of state-of-the-art reconstruction algorithms were also provided as a point of reference. Our aim is to facilitate the advancement of state-of-the-art CMR image reconstruction by introducing standardized evaluation criteria and making the dataset freely accessible to the research community. Researchers can access the dataset at https://***/#!Synapse:syn51471091/wiki/. Copyright © 2023, The Authors. All rights reserved.

关键词： Image reconstruction

Adaptive optics two-photon microscopy for in vivo imaging of mouse retina 6

学校读者我要写书评

暂无评论

Adaptive optics two-photon microscopy for in vivo imaging of...

Adaptive Optics and Wavefront Control for Biological Systems VI 2020

作者： Qin, Zhongya He, Sicong Chen, Congping Yang, Chao Yung, Jasmine Wu, Wanjie Leung, Christopher K. Liu, Kai Qu, Jianan Y. Biophotonics Research Laboratory Department of Electronic and Computer Engineering Hong Kong University of Science and Technology Clear Water Bay Kowloon Hong Kong Center of Systems Biology and Human Health School of Science and Institute for Advanced Study Hong Kong University of Science and Technology Clear Water Bay Kowloon Hong Kong Division of Life Science Hong Kong University of Science and Technology Clear Water Bay Kowloon Hong Kong Department of Ophthalmology and Visual Sciences Chinese University of Hong Kong Kowloon Hong Kong

ISBN: (纸本)9781510632592

Non-invasive retinal imaging has greatly facilitated the research of eye disease and neurodegenerative disorders in the central nervous system (CNS). Two-photon microscopy is a powerful tool for in vivo imaging of mouse retina because it provides intrinsic optical sectioning capability and the infrared laser is less likely to excite the photoreceptors. However, the dilated mouse eye has large optical aberrations, which must be corrected to achieve high-resolution or even diffraction-limited imaging. Here, we developed an adaptive optics (AO) two-photon microscope for in vivo imaging of retinal neurons through the eyeball of living mouse. We used the two-photon excited fluorescence signal of retina as the guide star to measure and correct the aberration of mouse eye. After AO correction, the fluorescence signal was increased by at least fivefold and the fine structures such as axons of retinal ganglion cells (RGC) were clearly resolved. To take advantage of the non-invasive high-resolution imaging, we demonstrated functional calcium imaging of RGC responding to the light stimulations. © 2020 SPIE.

关键词： Adaptive optics

One for Multiple: Physics-informed Synthetic Data Boosts Generalizable Deep Learning for Fast MRI Reconstruction

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Wang, Zi Yu, Xiaotong Wang, Chengyan Chen, Weibo Wang, Jiazheng Chu, Ying-Hua Sun, Hongwei Li, Rushuai Li, Peiyong Yang, Fan Han, Haiwei Kang, Taishan Lin, Jianzhong Yang, Chen Chang, Shufu Shi, Zhang Hua, Sha Li, Yan Hu, Juan Zhu, Liuhong Zhou, Jianjun Lin, Meijing Guo, Jiefeng Cai, Congbo Chen, Zhong Guo, Di Yang, Guang Qu, Xiaobo Department of Electronic Science Intelligent Medical Imaging R&D Center Fujian Provincial Key Laboratory of Plasma and Magnetic Resonance National Institute for Data Science in Health and Medicine Xiamen University China Human Phenome Institute Fudan University China Philips Healthcare China Siemens Healthineers Ltd. China United Imaging Research Institute of Intelligent Imaging China Department of Nuclear Medicine Nanjing First Hospital China Shandong Aoxin Medical Technology Company China Department of Radiology The First Affiliated Hospital of Xiamen University China Department of Radiology Zhongshan Hospital Affiliated to Xiamen University China China Department of Cardiology Shanghai Institute of Cardiovascular Diseases Zhongshan Hospital Fudan University China Department of Radiology Zhongshan Hospital Fudan University China Department of Cardiovascular Medicine Heart Failure Center Ruijin Hospital Lu Wan Branch Shanghai Jiaotong University School of Medicine China Department of Radiology Ruijin Hospital Shanghai Jiaotong University School of Medicine China Medical Imaging Department The First Affiliated Hospital of Kunming Medical University China Xiamen Key Laboratory of Clinical Transformation of Imaging Big Data and Artificial Intelligence China Department of Applied Marine Physics and Engineering Xiamen University China Department of Microelectronics and Integrated Circuit Xiamen University China School of Computer and Information Engineering Xiamen University of Technology China Department of Bioengineering Imperial College London United Kingdom

Magnetic resonance imaging (MRI) is a widely used radiological modality renowned for its radiation-free, comprehensive insights into the human body, facilitating medical diagnoses. However, the drawback of prolonged scan times hinders its accessibility. The k-space undersampling offers a solution, yet the resultant artifacts necessitate meticulous removal during image reconstruction. Although Deep Learning (DL) has proven effective for fast MRI image reconstruction, its broader applicability across various imaging scenarios has been constrained. Challenges include the high cost and privacy restrictions associated with acquiring large-scale, diverse training data, coupled with the inherent difficulty of addressing mismatches between training and target data in existing DL methodologies. Here, we present a novel Physics-Informed Synthetic data learning framework for Fast MRI, called PISF. PISF marks a breakthrough by enabling generalized DL for multi-scenario MRI reconstruction through a single trained model. Our approach separates the reconstruction of a 2D image into many 1D basic problems, commencing with 1D data synthesis to facilitate generalization. We demonstrate that training DL models on synthetic data, coupled with enhanced learning techniques, yields in vivo MRI reconstructions comparable to or surpassing those of models trained on matched realistic datasets, reducing the reliance on real-world MRI data by up to 96%. Additionally, PISF exhibits remarkable generalizability across multiple vendors and imaging centers. Its adaptability to diverse patient populations has been validated through evaluations by ten experienced medical professionals. PISF presents a feasible and cost-effective way to significantly boost the widespread adoption of DL in various fast MRI applications. Importantly, it alleviates the ethical and practical challenges associated with in vivo human data acquisitions, opening new horizons for advancing fast MRI technologies. Copyright © 20

关键词： Magnetic resonance imaging

Dispensed transformer network for unsupervised domain adaptation

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Li, Yunxiang Li, Jingxiong Dan, Ruilong Wang, Shuai Jin, Kai Zeng, Guodong Wang, Jun Pan, Xiangji Zhang, Qianni Zhou, Huiyu Jin, Qun Wang, Li Wang, Yaqi College of Computer Science and Technology Hangzhou Dianzi University Hangzhou China Artificial Intelligence and Biomedical Image Analysis Lab Westlake University Hangzhou China School of Mechanical Electrical and Information Engineering Shandong University Weihai China Department of Ophthalmology Hospital of Zhejiang University Hangzhou China sitem Center for Translational Medicine and Biomedical Entrepreneurship University of Bern Bern Switzerland School of Biomedical Engineering Shanghai Jiao Tong University Shanghai China School of Electronic Engineering and Computer Science Queen Mary University of London London United Kingdom School of Computing and Mathematical Sciences University of Leicester United Kingdom Department of Human Informatics and Cognitive Sciences Faculty of Human Sciences Waseda University Tokyo Japan Department of Radiology and Biomedical Research Imaging Center University of North Carolina at Chapel Hill Chapel Hill United States College of Media Engineering Communication University of Zhejiang Hangzhou China

Accurate segmentation is a crucial step in medical image analysis and applying supervised machine learning to segment the organs or lesions has been substantiated effective. However, it is costly to perform data annotation that provides ground truth labels for training the supervised algorithms, and the high variance of data that comes from different domains tends to severely degrade system performance over cross-site or cross-modality datasets. To mitigate this problem, a novel unsupervised domain adaptation (UDA) method named dispensed Transformer network (DTNet) is introduced in this paper. Our novel DTNet contains three modules. First, a dispensed residual transformer block is designed, which realizes global attention by dispensed interleaving operation and deals with the excessive computational cost and GPU memory usage of the Transformer. Second, a multi-scale consistency regularization is proposed to alleviate the loss of details in the low-resolution output for better feature alignment. Finally, a feature ranking discriminator is introduced to automatically assign different weights to domain-gap features to lessen the feature distribution distance, reducing the performance shift of two domains. The proposed method is evaluated on large fluorescein angiography (FA) retinal nonperfusion (RNP) cross-site dataset with 676 images and a wide used cross-modality dataset from the MM-WHS challenge. Extensive results demonstrate that our proposed network achieves the best performance in comparison with several state-of-the-art techniques. © 2021, CC BY.

关键词： Supervised learning