Smart Imaging Systems

Organizers
Saiprasad Ravishankar, University of Michigan, USA
Greg Ongie, University of Michigan, USA
Jeff Fessler, University of Michigan, USA 

Speakers

  1. Jong Chul Ye, Korea Advanced Institute of Science & Technology, Republic of Korea
  2. Jeffrey Fessler, University of Michigan, USA
  3. Rebecca Willett, University of Wisconsin-Madison, USA
  4. Michal Sofka, 4Catalyzer and Buttery Network, New York, USA
  5. Volkan Cevher, Swiss Federal Institute of Technology Lausanne, Switzerland
  6. Leslie Ying, University at Buffalo – SUNY, USA

Abstract

Data-driven techniques have received increasing attention in recent years for solving various problems in biomedical imaging. Data-driven models and approaches such as dictionary or transform learning, deep learning, etc., provide promising performance in image reconstruction problems in magnetic resonance imaging, computed tomography, and other modalities relative to traditional approaches using hand-crafted models such as the discrete cosine transform, wavelets, or total variation. The focus of this proposed session is on “smart imaging systems” – this term encompasses the latest approaches for making all components (either individually or jointly) of an imaging system data-driven, including data acquisition and sampling, image reconstruction, and processing/analytics. Smart imaging systems would continually learn from big datasets and on-the-fly and adapt themselves for speed, efficiency, and image performance or quality.

The special session will focus more on the signal processing and learning aspects of such smart systems rather than new hardware developments. The approaches presented in invited talks will exploit imaging physics, sophisticated measurement and image models and image reconstruction models including tensor and manifold structures, graphical models, deep learning, etc., along with advanced learning and optimization techniques for incorporating the models in real world imaging applications. Various imaging modalities and applications will be covered, along with theoretical analysis and understanding of the data-driven approaches. The session will bring together experts in academia and industry working in biomedical imaging, machine learning, signal processing, compressed sensing, optimization, and related areas, and facilitate substantive and cross-disciplinary interactions on cutting-edge smart imaging methods and systems.

Other Special Sessions