Evaluation of imaging biomarkers for identification of single cancer cells in blood

Masao Odaka, Hyonchol Kim, Mathias Girault, Akihiro Hattori, Hideyuki Terazono, Kenji Matsuura, Kenji Yasuda*

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract

A method of discriminating single cancer cells from whole blood cells based on their morphological visual characteristics (i.e., "imaging biomarker") was examined. Cells in healthy rat blood, a cancer cell line (MAT-LyLu), and cells in cancer-cell-implanted rat blood were chosen as models, and their bright-field (BF, whole-cell morphology) and fluorescence (FL, nucleus morphology) images were taken by an on-chip multi-imaging flow cytometry system and compared. Eight imaging biomarker indices, i.e., cellular area in a BF image, nucleus area in an FL image, area ratio of a whole cell and its nucleus, distance of the mass center between a whole cell and nucleus, cellular and nucleus perimeter, and perimeter ratios were calculated and analyzed using the BF and FL images taken. Results show that cancer cells can be clearly distinguished from healthy blood cells using correlation diagrams for cellular and nucleus areas as two different categories. Moreover, a portion of cancer cells showed a low nucleus perimeter ratio less than 0.9 because of the irregular nucleus morphologies of cancer cells. These results indicate that the measurements of imaging biomarkers are practically applicable to identifying cancer cells in blood.

Original languageEnglish
Article number06FN04
JournalJapanese journal of applied physics
Volume54
Issue number6
DOIs
Publication statusPublished - 2015 Jun 1
Externally publishedYes

ASJC Scopus subject areas

  • Engineering(all)
  • Physics and Astronomy(all)

Fingerprint

Dive into the research topics of 'Evaluation of imaging biomarkers for identification of single cancer cells in blood'. Together they form a unique fingerprint.

Cite this