A model of the spatial microenvironment of the colonic crypt

There have been great advancements in the field of immunofluorescence imaging. The surge in development of analytical methods for such data makes it crucial to develop benchmark synthetic datasets for objectively validating these methods. We propose a model of the healthy colonic crypt microenvironments. Our model can simulate immunofluorescence image data with parameters that allow control over cellularity, cell overlap ratio, image resolution, and objective level. To the best of our knowledge, ours is the first model to simulate immunofluorescence image data at subcellular level for healthy colon tissue, where the cells have several compartments and are organized to mimic the microenvironment of tissue in situ rather than dispersed cells in a cultured environment. Validation of the model has been performed by comparing morphological features of the tissue structure between real and simulated images. In addition, we compare the performance of two cell counting algorithms. The simulated data could also be used to validate techniques such as image restoration, cell segmentation, and crypt segmentation.