The MICCAI community needs data with known ground truth to develop, evaluate, and validate computerised image analytic tools, as well as to facilitate clinical training. Since synthetic data are ideally suited for this purpose, over the years, a full range of models underpinning image simulation and synthesis have been developed: (a) machine and deep learning methods based on generative models, (b) simplified mathematical models to test segmentation, tracking, restoration, and registration algorithms; (c) detailed mechanistic models (top–down), which incorporate priors on the geometry and physics of image acquisition and formation processes; and (d) complex spatio-temporal computational models of anatomical variability, organ physiology, morphological changes in tissues or disease progression.
The goal of the Simulation and Synthesis in Medical Imaging (SASHIMI) workshop is to bring together all those interested in such problems for invigorating research, discussing current approaches, and stimulating new ideas and scientific directions in this field. The objectives are to: (a) hear from invited speakers in the areas of transfer learning, generative adversarial networks, variational auto encoders, and biophysical models, and cross-fertilize across these fields; (b) bring together experts of image synthesis to raise the state of the art; and (c) identify challenges and opportunities for further research. We also would like to identify the suitable approaches to evaluate the plausibility of synthetic data and to collect benchmark data that can help with the development of future algorithms.
Topics of interest include, but are not limited to, the following: