VALSE Webinar 21-28期 总第256期 Transformer在医学图像处理的应用

报告时间	2021年10月20日 (星期三) 晚上20:00 (北京时间)
主  题	Transformer在医学图像处理的应用
主持人	雷柏英 (深圳大学)
直播地址	https://live.bilibili.com/22300737

报告嘉宾：付华柱 (IHPC, A*STAR)

报告题目：Transformers 在多模态 MR 成像中的应用

报告嘉宾：戈宗元 (Monash University)

报告题目：Transformer在皮肤病分层诊断和癫痫药物推荐上的应用

报告嘉宾：于乐全 (香港大学)

报告题目：Transformer在医学图像分析和重建中的应用

Panel嘉宾：

付华柱 (IHPC, A*STAR)、戈宗元 (Monash University)、于乐全 (香港大学)、刘明霞 (北卡罗来纳大学教堂山分校)、周泸萍 (悉尼大学)、秦璟 (香港理工大学)、夏勇 (西北工业大学)

Panel议题：

1. 利用Transformer训练网络的初始阶段，模型往往并不稳定，请问有什么好的解决办法？

2. 一般来说，Transformer对大体量数据集的效果比较好，但对医学图像而言，数据量普遍较少，那么如何能充分利用Transformer的自注意力的优势来解决医学图像数据量小的短板呢？

3. 有什么比较有效的方法解决Transformer在图像（特别是医学图像尺寸较大和三维医学影像）中计算量较大的问题？

4. 怎么选择Transformer、CNN、MLP，三者的优势分别是什么，有没有最新的方法将三者的优点结合起来？

5. Transformer强大的建模能力是否可以在计算机视觉中大放异彩，今后的研究思路又有哪些？

6. Transformer在计算机视觉中的限制和挑战是什么？哪样的结构设计更高效灵活？

*欢迎大家在下方留言提出主题相关问题，主持人和panel嘉宾会从中选择若干热度高的问题加入panel议题！

报告嘉宾：付华柱 (IHPC, A*STAR)

报告时间：2021年10月20日 (星期三)晚上20:00 (北京时间)

报告题目：Transformers 在多模态 MR 成像中的应用

报告人简介：

Huazhu Fu is a senior scientist at Institute of High Performance Computing (IHPC), Agency for Science, Technology and Research (A*STAR). He received his Ph.D. from Tianjin University in 2013 and was a Research Fellow at NTU for two years. From 2015 to 2018, he was a research scientist in I2R at A*STAR, Singapore. From 2018 to 2021, he was a senior scientist in Inception Institute of Artificial Intelligence, UAE. His research interests include computer vision, machine learning, and AI in health. He has published more than 120 publications, which have been cited over 6600 times according to Google Scholar. As a co-author, he is a recipient of ICME 2021 Best Paper Award. He currently serves as an Associate Editor of IEEE TMI, and IEEE JBHI. He also serves as an AC/SPC of MICCAI 2021, and AAAI 2022, and co-chair of OMIA workshop.

个人主页：

https://hzfu.github.io/

报告摘要：

Accelerated multi-modal magnetic resonance (MR) imaging is a new and effective solution for fast MR imaging, providing superior performance in restoring the target modality from its undersampled counterpart with guidance from an auxiliary modality. In this talk, I will introduce our proposed multimodal transformer (MTrans), which is capable of transferring multi-scale features from the target modality to the auxiliary modality, for accelerated MR imaging. Our framework provides three appealing benefits: (i) Our MTrans is the first attempt at using improved transformers for multimodal MR imaging, affording more global information compared with existing CNN-based methods. (ii) A new cross attention module is proposed to exploit the useful information in each modality at different scales. It affords both distinct structural information and subtle pixel-level information, which supplement the target modality effectively. (iii) We evaluate MTrans with various accelerated multimodal MR imaging tasks, e.g., MR image reconstruction and super-resolution, where MTrans outperforms state-of-the-art methods on fastMRI and real-world clinical datasets.

参考文献：

[1] "Accelerated Multi-Modal MR Imaging with Transformers",Chun-Mei Feng, Yunlu Yan, Geng Chen, Huazhu Fu, Yong Xu, and Ling Shao,arXiv, arXiv:2106.14248.

[2] "Multi-Contrast MRI Super-Resolution via a Multi-Stage Integration Network",Chun-Mei Feng, Huazhu Fu, Shuhao Yuan, and Yong Xu,MICCAI, 2021.

[3] "Task Transformer Network for Joint MRI Reconstruction and Super-Resolution", Chun-Mei Feng, Yunlu Yan, Huazhu Fu, Li Chen, and Yong Xu, MICCAI, 2021.

报告嘉宾：戈宗元 (Monash University)

报告时间：2021年10月20日 (星期三)晚上20:30 (北京时间)

报告题目：Transformer在皮肤病分层诊断和癫痫药物推荐上的应用

报告人简介：

戈宗元博士是澳大利亚莫纳什大学工程院和VC office的副教授，Monash Medical AI研究中心 (https://www.monash.edu/mmai-group)的主任，主要研究领域为统计分析，机器学习，医学人工智能以及流行病学。截止目前，他作为一作和重要作者在The Lancet Digital Health, The British Medical Journal, Bioinformatics, Hypertension, IEEE Transactions on Pattern Analysis and Machine Intelligence, IEEE Transactions on Medical Imaging, NeurIPS, CVPR, ICCV, ICLR, KDD, AAAI and MICCAI等期刊和会议已发表90多篇同行评审的出版物和专利，仅博士毕业四年，他的google scholar引用超过3000+,作为PCI和CI所获得的研究资助超过1500万澳元。

戈教授在皮肤病学和Molemap，IBM,澳大利亚皮肤癌研究中心 (ACEMID)合作产生了10篇医学科研论文和2项专利，三项项目正在进行临床试验，他的这项工作在2017年被授予IBM内部最高科学奖和年度员工，这项研究被用于皮肤癌防治并且已经服务了几十万的大众。除此之外，他也是Airdoc，鹰瞳科技研究团队的首席科学家。他参与研发了基于千万数据训练的全眼底检查平台。这个系统在眼底糖网，老年黄斑病变和青光眼防治做出了卓越贡献。截止2021年，Airdoc的眼底扫描平台已经在院内和院外服务了多达千万级别的民众。

个人主页：

https://zongyuange.github.io/

报告摘要：

Transformer ("Attention is all you need"), as a trend in the artificial intelligence community in recent years, has achieved excellent results on various computer vision tasks. In this presentation, I will mainly discuss our group's application prospects and results using Transformer in hierarchical skin classification, clustering learning, and AI-powered epilepsy drug selection.

参考文献：

[1] Z.Chen, B.Rollo, A.Baker, A.Anderson, Y.Ma, T.Obrien, Z.Ge, X.Wang, P.Kwan, New era of personalised epilepsy management, The British Medical Journal (BMJ), 2020

[2] Y.Zhen, V.Mar, A.Eriksson, S.Chandra, P.Bonnington, L.Zhang, Z.Ge, End-toend ugly duckling sign detection for melanoma identification with transformers, International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI), 2021

[3] Chen, Z., Anderson, A., Z.Ge, and Kwan, P., One step closer towards personalized epilepsy management. Brain, 2021

[4] Lin, D., Xiong, J., Liu, C., Zhao, L., Li, Z., Yu, S., Wu, X., Z.Ge., Hu, X., Wang, B. and Fu, M., 2021. Application of Comprehensive Artificial intelligence Retinal Expert (CARE)system: a national real-world evidence study. The Lancet Digital Health, 2021.

报告嘉宾：于乐全 (香港大学)

报告时间：2021年10月20日 (星期三)晚上21:00 (北京时间)

报告题目：Transformer在医学图像分析和重建中的应用

报告人简介：

Dr. Lequan Yu is an Assistant Professor at the Department of Statistics and Actuarial Science, the University of Hong Kong. Before joining HKU, he was a postdoctoral fellow at Stanford University. He obtained his Ph.D. degree from The Chinese University of Hong Kong and bachelor’s degree from Zhejiang University, both in Computer Science. He also experienced research internships in Nvidia and Siemens Healthineers. His research interests are developing advanced machine learning methods for biomedical data analysis, with a primary focus on medical images. He has won the CUHK Young Scholars Thesis Award 2019 and has also won the Best Paper Awards of Medical Image Analysis-MICCAI in 2017 and International Workshop on Machine Learning in Medical Imaging in MICCAI 2017. He serves as the senior PC member of IJCAI, AAAI, and the reviewer for top-tier journals and conferences, such as Nature Machine Intelligence, IEEE-PAMI, IEEE-TMI, Medical Image Analysis, etc.

个人主页：

https://yulequan.github.io/

报告摘要：

Recently, transformers have drawn scant attention from the medical imaging community. Given the ability to exploit long-term dependencies, transformers are promising to help convolutional neural networks (convnets)to overcome its inherent shortcomings of spatial inductive bias. This talk will present our recent works in applying transformer for medical imaging problems, including 3D medical image segmentation and low dose computed tomography. We will also discuss some potential directions of transformer for medical imaging.

参考文献：

[1] Zhicheng Zhang, Lequan Yu, Xiaokun Liang, Wei Zhao, and Lei Xing. "TransCT: Dual-path Transformer for Low Dose Computed Tomography." Medical Image Computing and Computer Assisted Intervention (MICCAI), 2021.

[2] Hong-Yu Zhou, Jiansen Guo, Yinghao Zhang, Lequan Yu, Liansheng Wang, and Yizhou Yu. "nnFormer: Interleaved Transformer for Volumetric Segmentation." arXiv preprint arXiv:2109.03201 (2021).

Panel嘉宾：刘明霞 (北卡罗来纳大学教堂山分校)

嘉宾简介：

Dr. Liu is an Assistant Professor of the Department of Radiology and Biomedical Research Imaging Center (BRIC)at the University of North Carolina at Chapel Hill. She has a solid background in machine learning and medical image analysis, leading a wide range of research spanning biomedical image reconstruction, processing, analysis, and neuroscience. Her primary research interest lies in studying pathological changes in the human brain by creating innovative machine learning technologies for automated brain disease analysis. Current efforts are being directed to extending these works to harmonize/adapt multi-site data, integrate multiple data modalities (e.g., MRI, PET, fMRI, DTI, CSF, and genetics data), and apply the resulting analysis paradigms to a number of important clinical studies, including Alzheimer’s disease, mild cognitive impairment (MCI), preclinical AD (e.g., subjective cognitive decline, SCD), and major depressive disorders (MDD).

个人主页：

https://mingxia.web.unc.edu/

Panel嘉宾：周泸萍 (悉尼大学)

嘉宾简介：

Dr. Luping Zhou is a faculty member in School of Electrical and Information Engineering, The University of Sydney. She obtained her PhD from Australian National University and got her post-doctoral training in University of North Carolina at Chapel Hill. Dr. Zhou works on the interface of medical image analysis, machine learning, and computer vision, and has published 100+ research papers in these fields. Her current research is focused on medical image analysis with statistical graphical models and deep learning, as well as general visual recognition problems. She was a recipient of the prestigious ARC (Australian Research Council) DECRA award (Discovery Early Career Researcher Award). Dr. Zhou is the Associate Editor of the journals IEEE Trans. on Medical Imaging (TMI), Pattern Recognition and Neurocomputing. She is a Senior Member of IEEE.

个人主页：

https://sites.google.com/view/lupingzhou

Panel嘉宾：秦璟 (香港理工大学)

嘉宾简介：

QIN,Jing (Harry)is currently an associate professor in Centre for Smart Health, School of Nursing, The Hong Kong Polytechnic University. His research focuses on creatively leveraging advanced virtual reality (VR)and artificial intelligence (AI)techniques in healthcare and medicine applications and his achievements in relevant areas has been well recognized by the academic community. He won the Hong Kong Medical and Health Device Industries Association Student Research Award for his PhD study on VR-based simulation systems for surgical training and planning. He won 5 best paper awards for his research on AI-driven medical image analysis and computer-assisted surgery. He served as a local organization chair for MICCAI 2019, technical program committee (TPC)members for many academic conferences, speakers for many invited talks, and referees for many prestigious journals in relevant fields.

个人主页：

https://www.polyu.edu.hk/sn/people/academic-staff/dr-harry-qin/

Panel嘉宾：夏勇 (西北工业大学)

嘉宾简介：

夏勇，西北工业大学计算机学院教授。研究方向为医学影像智能计算。近三年，在IEEE-TPAMI/TMI/TIP/JBHI、PR、NeurIPS、CVPR、IJCAI、MICCAI发表论文40余篇，并在ISBI 2019 C-NMC、PROMISE12、BraTS2020、MyoPS 2020、COVID-19-20等竞赛获得优异成绩；现为中国图象图形学学会视觉大数据专委会常委、中国抗癌协会肿瘤影像专业委员会人工智能学组副组长、陕西省计算机学会人工智能专委会主任；先后担任ISBI 2017分会主席、MICCAI 2019地区主席和MICCAI 2020分会主席等。

个人主页：

https://teacher.nwpu.edu.cn/yongxia.html

主持人：雷柏英 (深圳大学)

主持人简介：

雷柏英，教育部青年长江学者入选者，(留学)博士生导师，深圳大学研究员、深圳大学“荔园优青”、深圳市海外高层次人才 (孔雀计划)、深圳市高层次后备级人才，深圳市孔雀团队核心成员等，在新加坡南洋大学博士学位。先后在美国北卡大学教堂山分校和法国计算和自动化研究所等进行研究和访问。主要研究医学图像处理和人工智能。在IEEE TMI、IEEE TNNLS、MedIA等以第一/通讯作者 (含共同)等发表SCI论文95篇 (18篇IEEE汇刊、28篇中科院1区、1篇ESI高被引)，共发表了200余篇国际期刊和会议论文。谷歌学术总引用3571次，单篇文章最高引用232次，H指数29。获授权专利18项 (3项已转让)。主持国家自然科学基金面上等项目共18项。现任IEEE TNNLS副主编 (Associate Editor)、IEEE TMI 副主编 (Associate Editor)、Medical Image Analysis (中国4人)编委、 Neural Comp uting & Application 等6种SCI期刊编委。现为IEEE 高级会员，IEEE Bio Imaging Signal Processing (BISP)Technical Committee (TC)技术委员会委员 (中国1人)， Biomedical Imaging and Image Processing (BIIP)Technical Committee (TC)技术委员会委员 (中国2人)，医学图像顶级学术会议MICCAI2021领域主席，生物医学工程青工委委员。指导学生获MICCAI国际竞赛3项任务冠军。获吴文俊人工智能科学技术奖三等奖 (排名第3)，深圳市科学技术奖一等奖 (排名第3)。

个人主页：

http://bme.szu.edu.cn/20181/0612/66.html

特别鸣谢本次Webinar主要组织者：

主办AC：雷柏英 (深圳大学)

活动参与方式

1、VALSE每周举行的Webinar活动依托B站直播平台进行，欢迎在B站搜索VALSE_Webinar关注我们！

直播地址：

https://live.bilibili.com/22300737；

历史视频观看地址：

https://space.bilibili.com/562085182/ 

2、VALSE Webinar活动通常每周三晚上20:00进行，但偶尔会因为讲者时区问题略有调整，为方便您参加活动，请关注VALSE微信公众号：valse_wechat 或加入VALSE QQ R群，群号：137634472）；

*注：申请加入VALSE QQ群时需验证姓名、单位和身份，缺一不可。入群后，请实名，姓名身份单位。身份：学校及科研单位人员T；企业研发I；博士D；硕士M。

3. VALSE微信公众号一般会在每周四发布下一周Webinar报告的通知。

4、您也可以通过访问VALSE主页：http://valser.org/ 直接查看Webinar活动信息。Webinar报告的PPT（经讲者允许后），会在VALSE官网每期报告通知的最下方更新。

看到这了，确定不关注下吗？关注下再走吧～