Seminar Machine Learning and Health

Change of room! Room H0.08 is behind (South-East of) the main building 904. It is opposite the parking P7. See map below for the H-building.

Program:

Prof. Bram van Ginneken (http://www.diagnijmegen.nl/index.php/Bram_van_Ginneken)(Radboud UMC - Diagnostic Image Analysis Group) - Interpreting Medical Images: Humans versus Computers

Prof. Paul van Diest (http://www.cgdb.nl/research-groups/28/) (UMC Utrecht - Pathologie) - Image analysis applications on pathology digital slides

dr. Vlado Menkovski (Philips) - Training and understanding the performance of Deep Neural Networks on medical image data

dr. Guido van Wingen (https://www.spinozacentre.nl/person/43)(AMC - Spinoza Centre for Neuroimaging) - Machine learning for clinical neuroscience

Moderator: prof. Max Welling (http://www.uva.nl/en/about-the-uva/organisation/staff-members/content/w/e/m.welling/m.welling.html) (UvA - AMLab - A'dam Machine Learning Lab)

Abstracts:

Prof. Bram van Ginneken (http://www.diagnijmegen.nl/index.php/Bram_van_Ginneken)(Radboud UMC - Diagnostic Image Analysis Group) - Interpreting Medical Images: Humans versus Computers

Abstract: The first attempts to let computers make diagnosis based on medical image date back half a century. In this talk I will show various approaches to computer-aided diagnosis, discuss the potential for computers to outperform human readers in various tasks, and present recent advances based on convolution neural networks.

Bram van Ginneken is Professor of Functional Image Analysis at Radboud University Medical Center and co-chair of the Diagnostic Image Analysis Group within the Department of Radiology and Nuclear Medicine. He also works for Fraunhofer MEVIS in Bremen, Germany, and is one of the founders of Thirona, a company that provides quantitative analysis of chest CTscans. Bram studied Physics at the Eindhoven University of Technology and at the Utrecht University, where he obtained his Ph.D. at the Image Sciences Institute (ISI) on Computer-Aided Diagnosis in Chest Radiography.

Prof. Paul van Diest (http://www.cgdb.nl/research-groups/28/) (UMC Utrecht - Pathologie) - Image analysis applications on pathology digital slides

Abstract: Implementation of technology to fully scan histological slides and view them on a computer screen is becoming more and more popular in pathology as an alternative for traditional microscopy. This so-called ‘digital pathology’ has many advantages including the option to apply image analysis algorithms to objectify and support pathology diagnosis. In this presentation, an overview will be given of the potential and current status of image analysis applications of pathology digital slides.

Paul J van Diest studied Medicine, did his PhD and pathology residency at VU University Medical Center in Amsterdam; after obtaining his Board certification (1996) he became Consultant Pathologist, Associate Professor (1999) and full Professor (2001). Since 2003 he is Head of the Department of Pathology at University Medical Center Utrecht. He is Adjunct Professor of Oncology at the Sidney Kimmel Oncology Center at Johns Hopkins, Baltimore, USA, serves on the editorial board of international journals, and has been active in several international societies. He published >650 papers in peer reviewed journals, supervised 59 PhD theses; his H-index is 65.

dr. Vlado Menkovski (Philips) - Training and understanding the performance of Deep Neural Networks on medical image data

Abstract: Deep Learning methods have allowed for significant breakthroughs in image analysis. However, applied in the medical domain they face additional challenges. In this domain, it is common that annotated training data is scarce. Moreover, various pathologies for which accurate performance is particularly important are extremely rare. In this talk I will present example approaches that face these challenges by careful design of the neural network architecture, training procedure and exhaustive augmentation of the training data.

Vlado Menkovski is a Research Scientist in the Data Science department at Philips Research in Eindhoven. His research focuses on Machine Learning methods for medical image analysis, genomics and patient monitoring. Vlado studied Computer Science at the Eindhoven University of Technology, where he obtained his Ph.D. for which he won the annual best doctoral project award. Vlado is an author of more than 40 journal, conference and book chapter publications.

dr. Guido van Wingen (https://www.spinozacentre.nl/person/43)(AMC - Spinoza Centre for Neuroimaging) - Machine learning for clinical neuroscience

Abstract: Machine learning analysis of neuroimaging data provides statistical inference on the individual level. In this talk I will highlight how machine learning may solve clinical problems such as differential diagnosis, disease prognosis, and the prediction of treatment outcome, and thereby enable personalized treatment for brain disorders.

Guido van Wingen is Principal Investigator for Neuroimaging in Psychiatry at the Academic Medical Center Amsterdam. His research focuses on revealing the neural underpinnings of psychiatric disorders, the neural mechanisms underlying successful treatment, and predicting treatment outcome using neuroimaging data

Entrance free. Please RSVP.

http://photos1.meetupstatic.com/photos/event/7/d/0/e/600_446552014.jpeg

http://photos4.meetupstatic.com/photos/event/7/d/2/d/600_446552045.jpeg