Highly accurate segmentation of cell areas based on DIC images using deep learning
Quantification of cell migration and cell confluency is important in biological and medical research on cellular functions. A scratch assay is used to quantitatively measure the speed of cells migrating to a cell-free area (gap) that is physically produced. This is a common technique for evaluating cell migration in cell development and differentiation, as well as in the invasion and metastasis of cancer cells. However, manually processing unstained sample images for quantification takes an immense amount of time. In addition, the Wound Healing function, a special application for scratch assays in the NIS-Elements imaging software, shows roughly correct results, but it has limited accuracy in terms of detailed detection.
In this Application Note, we introduce examples of quantification of a scratch assay using the NIS.ai module of NIS-Elements. These examples proved that NIS.ai can make more accurate inferences compared to the existing Wound Healing function based on a small number of training images.