Deep learning in Medical Imaging

Special Session 4: Thursday, 14 April (4:00pm-5:30pm)

Organizer

Debdoot Sheet

Speakers

  • Ronald M. Summers (National Institutes of Health Clinical Center, Bethesda, MD, USA)
  • Hayit Greenspan (Tel Aviv University, Israel)
  • Olaf Ronneberger (Albert-Ludwigs-Universität Freiburg, Germany)

Keywords

classification, image segmentation, machine learning.

Abstract

Deep learning is steadily growing for solving biomedical and biological imaging and image analysis problems, with increased participation of industry and academia. While machine learning is traditionally the playfield for computer sciences, early adopters of deep learning from biomedical sciences face challenges due to limited exposure to its fundamental concepts. This special session is expected to introduce early and young adopters of this technique to the perspectives on: (i) notional hierarchical embedding in biomedical and biological images, and understanding hierarchical knowledge transfer in classical approaches, (ii) common practices in deep learning including tool-(box)-s of the trade (e.g. AEs, CNNs or RBMs), (iii) handling big-biomedical data for deep learning with limited memory and computation resources, (iv) using deep networks for feature discovery followed by ensemble learning for classification and (v) understanding and analyzing attributes learnt by a deep network, its locality and stability over more learning instances and some challenging questions yet to be answered.

Other special sessions

Big Data in Medical Imaging
Quantitative Musculoskeletal Imaging
Biomarker Detection and Discovery in Histopathology Images
Frontiers in Pulmonary Image Analysis
3D Image Analysis and Stereology in Fluorescence Microscopy
3D Echocardiography: Towards Ultrafast