Application of Support Vector Machine to Recognize Trans-differentiated Neural Progenitor Cells for Bright-Field Microscopy

One possible solution of the investigation of the cell fate decision and its function is the study of cell morphology. Bright-field imaging analysis allow us to use a labeling free and non-invasive approach to measure the morphological dynamics during cellular reprogramming, which includes induced pluripotent stem cells (iPSCs), and trans-differentiated neural progenitor cells (NPCs) from somatic cell source. In order to automatically analyze and cultivate cells, a system classifying NPCs under bright-field microscopic imaging is necessary. In this paper, we investigate the use of support vector machine (SVM) based on a set of features for this task. The results illustrate that such a data driven approach has remarkable recognition and generalization performance.

I. Introduction

Cellular reprogramming opens the door for personalized regenerative medicine especially in fight with chronic and degenerative disease. We have established a technology converting the cells from urine into neural progenitor cells (NPC), so called trans-differentiation. This technology can allow us easily obtain source cells in a noninvasive approach. However, further studies we found that, during the trans-differentiation process, along with positive NPCs, the negative colonies have different sub type of morphology and polarity. As we understand, the cellular polarity strongly linked with gene expression, cell cycle and other cellular regulation may explain the mechanism regarding the different route of reprogramming. For example cell polarity changes between MET and EMT which linked with induced pluripotent stem cell (iPSCs) colony formation or tumor genesis. Hence, the regulation of cell fate changes has a strong link with morphology as a read out.