Gelatin scaffolds that mimic the structure of the Lymph Nodes (LN) can assist in studying the influence of Fibroblastic Reticular Cells, supportive cells in the LN, on peripheral tolerance in Type I diabetes. We fabricated freeze-dried gelatin scaffolds crosslinked with glutaraldehyde using various experimental parameters like crosslinking time, gelatin concentration, and freezing conditions. The scaffolds were sputter-coated with a thin layer of gold and imaged by Scanning Electron Microscope (SEM) to assess porosity and pore size. We currently use ImageJ to quantify the average pore size of a sample by hand-outlining 20 pores from each scaffold's SEM image and measuring their Feret radii. To optimize this calculation process, we have adapted an existing MATLAB script by Rabbani et al. This script uses watershed segmentation to identify pores, then calculates pore radii using either Feret radii or a Circular approximation. For this study, we calculated the average pore radius with outliers, without outliers, or with filters applied to remove noisy values <20 μm and >200 μm using both methods. Average pore radii of four SEM images were found for each method. The code computing Feret radii was closer to ImageJ values for scaffolds with larger pores, whereas the Circular radii command was more similar to ImageJ for scaffolds with smaller pores. Further studies will identify common trends. In the future, we will examine the accuracy of a 20 pore ImageJ sample, identify more effective noise filters, and explore the impact of porosity on scaffold structure.
