Noninvasive discrimination between human normal and cancer cells by analysis of intracellular distribution of phase-shift data

Abstract

Aiming to establish a method for the noninvasive discrimination of cancer cells from normal cells in adherent culture, we investigated to employ all phase shift data for all pixels inside a cell. The bird's-eye views of phase shifts of human prostate epithelial cells (PRECs) and human prostatic carcinoma epithelial cell (PC-3) lines acquired by phase-shifting laser microscopy showed tableland and cone shapes, respectively, while treatment of PRECs with cytochalasin D resulted in the cone shape. So, the profile of phase shift in both sections towards the x- and y-axes of the views through the peaks of the phase shifts in PRECs and PC-3 cells were trapezoid-like and triangle-like, respectively. Typical profiles of phase shifts in a section in PRECs or PC-3 cells were calculated by averaging from 10 cells and smoothing. Cancer index is defined as the deduction of sums of the squared difference between a real cell and the typical profiles for a PREC and a PC-3 cell. The cancer indices for PC-3 and hepatocellular carcinoma cell lines were positive, while those for PRECs and human normal cryopreserved hepatocytes were negative. Cancer indices along the major axis of fibroblast-like cells of normal mesenchymal stem cells and the osteosarcoma cell line were negative and positive, respectively. Consequently, several cancer cells could be noninvasively discriminated from normal cells by calculating the cancer index employing phase shift for all pixels inside the cells.

Autoregressive parametric modeling combined ANOVA approach for label-free-based cancerous and normal cells discrimination

“…Artificial intelligence and deep learning techniques and method have proven to show an outstanding performance and capabilities in resolving recognition and classification problems. A combined deep learning approach in conjunction with expert trained data system was very powerful tool to identify cancer from gene expression data [ 45 ]. Moreover, it can also contribute towards the understanding the complex nature of cancer based on large public data as well.…”