Development of Cell Analysis Software to Evaluate Fibroblastic Changes in Cultivated Corneal Endothelial Cells for Quality Control.

Purpose: To develop software to evaluate the fibroblastic morphological changes in cultured human corneal endothelial cells (HCECs) as a quality control measure for use in tissue engineering therapy. Methods: Software was designed to recognize cell borders, to approximate cell shape as an ellipse, and to calculate the aspect ratio of the ellipse as an indicator of severity of the fibroblastic morphological change. Using the designed software, 60 phase contrast images of polygonal HCECs and 60 phase contrast images of fibroblastic HCECs were analyzed. The correlations of the aspect ratio and other parameters (cell density, percentage of cells surrounded by 6 cells, and coefficient of variation) were evaluated. Results: Cell shapes were recognized based on phase contrast images and were approximated as ellipses by software. The average aspect ratio was significantly higher (34.9%+/- 6.1%) in fibroblastic HCECs than in polygonal HCECs (24.4%+/- 2.3%) (P<0.01). The aspect ratio showed a correlation with cell density, with the percentage of cells surrounded by 6 neighboring cells, and with the coefficient of variation (Pearson correlation coefficients, -0.84, -0.38, and 0.66, respectively). Conclusions: We propose that fibroblastic alteration of HCECs can be evaluated by the cell morphology based on the aspect ratio. Software developed in this study, which can analyze the frequency and severity of fibroblastic alteration, will be useful for nondestructive assessment of cells destined for use in cell-based therapy for corneal endothelial decompensation.