Prof. Andrew E. Teschendorff

Chinese Academy of Sciences, China

Andrew Teschendorff studied Mathematical Physics at the University of Edinburgh (1990-1995) under the supervision of Physics Nobel Laureate Peter Higgs. In 2000 he obtained a PhD in Theoretical Physics from Cambridge University. In 2003 he became a Senior Research Fellow in Statistical Cancer Genomics at the University of Cambridge. In 2008 he moved to the University College London (UCL) Cancer Institute to work in Statistical Cancer Epigenomics and where he was awarded the Heller Research Fellowship. He currently holds an appointment as a PI at the CAS Shanghai Institute for Nutrition and Health, formerly a joint CAS-Max-Planck Partner Institute for Computational Biology, and remains an Honorary Research Fellow at the UCL Cancer Institute. Besides Statistical Cancer Epigenomics, his other research interests include Cancer System-omics & Systems Biology and Network Physics. He is an Associate Editor for various journals, notably Genome Biology, and a reviewer and statistical advisor for journals including Nature, NEJM and Science. He is the recipient of the Tait Medal and Robert Schlapp Prize in Physics, the Jennings Prize, Cambridge-MIT Initiative and Isaac Newton Trust Awards, a Wellcome Trust VIP Award, a CAS Visiting Professorship and a CAS-Royal Society Newton Advanced Fellowship. He holds various patents on algorithms for cancer risk prediction and cell-type deconvolution.

Speech Title: "Network Physics Approaches for Single-cell Omics and Personalized Medicine"

Abstract: Single-cell omic data continue to advance our understanding of fundamental molecular biology and to offer new strategies for personalized medicine. In this talk, I will describe a number of recent computational methods we have developed which have led to novel insights into cellular priming, aging and cancer development. One method called ELVAR addresses the challenge of testing for differential abundance of cell-types, by using cell-attribute aware clustering on the single-cell manifold which improves the sensitivity to detect differential abundance of cell-types. An application to single-nucleus RNA-Seq data reveals increased stem-cell fractions in colonic polyps, paving the way for single-cell based personalized cancer risk prediction. In the second part of the talk, I will describe methods for inferring differentiation activity of transcription factors (SCIRA) and for detecting priming events in multipotent cell populations (DICE). In the context of aging, SCIRA reveals cell-type specific patterns of age-associated changes of transcription factors that transcend tissue-type. In the context of priming, I demonstrate how primed cells and the transcription factors implicating in their priming to specific downstream cell-fates are detectable using only transcriptomic data without the need for single-cell epigenomics (scATAC-Seq data).

 

Prof. Huiyu Zhou

University of Leicester, UK

Dr. Huiyu Zhou received a Bachelor of Engineering degree in Radio Technology from Huazhong University of Science and Technology of China, and a Master of Science degree in Biomedical Engineering from University of Dundee of United Kingdom, respectively. He was awarded a Doctor of Philosophy degree in Computer Vision from Heriot-Watt University, Edinburgh, United Kingdom. Dr. Zhou currently is a full Professor at School of Computing and Mathematical Sciences, University of Leicester, United Kingdom. He has published over 450 peer-reviewed papers in the field. He was the recipient of "CVIU 2012 Most Cited Paper Award", “MIUA 2020 Best Paper Award”, “ICPRAM 2016 Best Paper Award” and was nominated for “ICPRAM 2017 Best Student Paper Award” and "MBEC 2006 Nightingale Prize". His research work has been or is being supported by UK EPSRC, ESRC, AHRC, MRC, EU, Royal Society, Leverhulme Trust, Invest NI, Puffin Trust, Alzheimer’s Research UK, Invest NI and industry.

Speech Title: "Learning Uncertainty in Image Understanding"

Abstract: There are many questions to answer in image interpretation and understanding. Uncertainty in image analysis needs strong and powerful modelling tools to describe the objects in the images. Artificial intelligence (AI) plays a very important role in the design of a robust tool for image representation. Using some examples from his own work on uncertainty analysis, Prof. Zhou will explore how AI can stimulate new concepts or development of dealing with complicated problems and lead us to novel adventures through these applications.

 

 

 

Prof. Hua Wang

Colorado School of Mines, USA

Dr. Hua Wang is a Professor of the Department of Computer Science at Colorado School of Mines. He is also the Director of the Data Science Program at Colorado School of Mines. His current research interests include machine learning and data mining, as well as their applications in medical image computing, health informatics, bioinformatics, computer vision, additive manufacturing, and cheminformatics. He is developing efficient algorithms with nice theoretical guarantees to solve practical problems involving large scale data. He has published more than 100 academic papers in top journals including IEEE TMI, IEEE TNNLS, Bioinformatics and top conferences including ICML, NIPS, AAAI, IJCAI, ICDM, SDM, MICCAI, RECOMB, ISMB, ECCB, PSB, CVPR, ICCV, ECCV, RSS, SIGIR, ACL, ACM MM.

Speech Title: "Scaling Multi-Instance Support Vector Machine for Cancer Detection Using Large-Scale Histopathological Images"

Abstract: Breast cancer is a type of cancer that develops in breast tissue, and, after skin cancer, it is the most commonly diagnosed cancer in women all over the world. Given that an early diagnosis is imperative to prevent breast cancer progression, many machine learning models have automated the histopathological classification of the different types of carcinomas. However, many of them are not scalable to the large dataset. In this study, we propose the novel Primal-Dual Multi-Instance Support Vector Machine (pdMISVM) to determine which tissue segments in an image exhibit an indication of an abnormality. We also derive the efficient optimization approach for the proposed method by bypassing the quadratic programming and least-squares problems, which are commonly employed to optimize Support Vector Machine (SVM) models in multi-instance learning. The proposed method is scalable to large datasets, and it is computationally efficient. We applied our method to the public BreaKHis dataset and achieved promising prediction performance and scalability for histopathological classification.

Assoc. Prof. Jian Wu

Tsinghua University, China

Wu Jian received the B.S. degree in Biomedical Engineering from Tsinghua University in 1999, M.S. degrees and Ph.D. in Biomedical Engineering from Tsinghua University in 2004 respectively. After his postdoctoral work in the research center of biomedical engineering at graduate school at Shenzhen, Tsinghua university in 2006, he joined the research center of biomedical engineering at graduate school at Shenzhen, Tsinghua university as Assistant Professor until now. His research interests have spanned various areas of biomedical imaging, biomedical signals, minimally invasive interventional diagnosis and treatment. Dr. Wu has been developing novel methodologies in the fields of signal and medical image processing, and computer aided surgery. Lately, Dr. Wu focuses on the research of integrated diagnosis and treatment technology of cardiovascular system, liver, digestive system and orthopedics. Dr. Wu has published more than 100 papers and holds more than ten domestic patents.

Speech Title: "Study on Noninvasive and High-Density Precision Heart Lesion Localization Method"

Abstract: Ectopic excitation of the heart can trigger arrhythmia. At present, catheter ablation has been widely used in the treatment of arrhythmias, but it has limited effect on some arrhythmias. The inaccurate and incomplete location of the ectopic excitation points that trigger arrhythmia during the operation leads to the high recurrence rate after catheter ablation. Preoperative noninvasive localization of cardiac lesions is very attractive and challenging for the clinical treatment of arrhythmia. In this paper, the method of non-invasive acquisition of high-density body surface signals, localization, and visualization of ectopic focal points in the heart is studied, and a set of ectopic excitation non-invasive localization system for the diagnosis and treatment of arrhythmia is developed. Through more than ten cases of clinical testing, the results show that the method proposed in this paper has high accuracy.