|Date:||Friday, October 30, 2015|
|Time:||14:00 - 17:00, followed by drinks|
|Location:||Meeting room 3.5, SURF Utrecht, Hoog Overborch 4th floor (Hoog Catharijne, left of Albert Heijn)|
|Registration:||Deadline October 29; open to all, register here|
|Organizer:||NeuroInformatics.NL, program committee Rembrandt Bakker, Paul Tiesinga|
"High troughput data acquisition systems for neuroimaging have become very powerful and widely available. The potential for scientific breakthroughs is enormous, but there is a show stopper: the human in the image analysis loop."
This is what we wrote two years ago to announce a workshop that focused on how far we can get with fully automated analysis algorithms that turn terabytes of data into manageable bits and pieces. In the 2015 workshop we focus on recent trends in automated analysis techniques, with exciting new developments in deep learning (Max Welling) and convolutional neural networks (Pim Moeskops, Mohsen Ghafoorian). But we will also learn that what we had two years ago is mature and ready to be put into medical practice (Wiro Niessen), guiding doctors in their assessment of the patient's health (Hugo Kuijf). The same mature techniques are assessed for their ability to consistently predict biomarkers in MR-based images measured from a variety of machine vendors and protocol parameters (Rembrandt Bakker). That said, there are still many image analysis challenges for which automated analysis is still in its infancy (Adrienne Mendrik).
Walking directions from Utrecht Central Station