This year, I was part of the organizing committee for the workshop on Shape in Medical Imaging (ShapeMI), which was held in conjunction with the conference on Medical Image Computing and Computer Assisted Interventions (MICCAI 2018) in Granada, Spain. ShapeMI provided a venue for researchers who work in shape/geometric modeling, learning, analysis, statistics, classification and applications to share ideas, present recent research results and interact with each other. The workshop was attended by more than 80 people. It included 19 poster presentations and 11 oral presentations, which were organized in three blocks: Shape Applications / Validation / Software, Shape Methods and Shape classification and deep learning.

Each block was preceded by a keynote speaker: Dr. Stanley Durleman, Dr. Michael Bronstein and Dr. Daniel Rueckert, respectively. As the Shape Software keynote speaker, Dr. Durleman discussed his software package, Deformetrica. To introduce the Shape Methods block, Dr. Bronstein presented several different approaches for representing shapes using neural networks. Finally, Dr. Rueckert presented his work on deep learning in cardiac imaging using shape priors as the keynote speaker for the Shape classification and deep learning block.

The workshop also included a Poster and Demo Session. During the session, Dr. Jared Vicory presented SlicerSALT, which had its first release the week before the workshop.

Kitware sponsored the best paper award, which went to Kris Campbell and Tom Fletcher from the Scientific Computing Institute at the University of Utah for their work, “Nonparametric Aggregation of Geodesic Trends for Longitudinal Data Analysis.”

Both the organizers and the attendees were excited about the lively discussions and great scientific content of the workshop. We look forward to repeating the success of ShapeMI at future MICCAI conferences.

Leave a Reply