Label-free classification of cells based on supervised machine learning of subcellular structures

• DOI: 10.1371/journal.pone.0211347
• Publication Date: 2019-01-29T19:08:20Z
• Authors: Yusuke Ozaki; Hidenao Yamada; Hirotoshi Kikuchi; Amane Hirotsu; Tomohiro Murakami; Tomohiro Matsumoto; Toshiki Kawabata; Yoshihiro Hiramatsu; Kinji Kamiya; Toyohiko Yamauchi; Kentaro Goto; Yukio Ueda; Shigetoshi Okazaki; Masatoshi Kitagawa; Hiroya Takeuchi; Hiroyuki Konno
• Keywords: Cell type

It is demonstrated that cells can be classified by pattern recognition of the subcellular structure non-stained live cells, and was performed machine learning. Human white blood five types cancer cell lines were imaged quantitative phase microscopy, which provides morphological information without staining quantitatively in terms optical thickness cells. Subcellular features then extracted from obtained images as training data sets for The built classifier successfully WBCs (area under ROC curve = 0.996). This label-free, non-cytotoxic classification based on QPM has potential to serve an automated diagnosis single