• Hiroyuki Takemura, Tomohiko Ai, Konobu Kimura, Kaori Nagasaka, Toshihiro Takahashi, Koji Tsuchiya, Haeun Yang, Aya Konishi, Kinya Uchihashi, Takashi Horii, Yoko Tabe, and Akimichi Ohsaka.
• Department of Clinical Laboratory, Juntendo University Hospital, Tokyo, Japan.
• Plos One. 2018 Jan 1; 13 (4): e0195923.

AbstractThe XN series automated hematology analyzer has been equipped with a body fluid (BF) mode to count and differentiate leukocytes in BF samples including cerebrospinal fluid (CSF). However, its diagnostic accuracy is not reliable for CSF samples with low cell concentration at the border between normal and pathologic level. To overcome this limitation, a new flow cytometry-based technology, termed "high sensitive analysis (hsA) mode," has been developed. In addition, the XN series analyzer has been equipped with the automated digital cell imaging analyzer DI-60 to classify cell morphology including normal leukocytes differential and abnormal malignant cells detection. Using various BF samples, we evaluated the performance of the XN-hsA mode and DI-60 compared to manual microscopic examination. The reproducibility of the XN-hsA mode showed good results in samples with low cell densities (coefficient of variation; % CV: 7.8% for 6 cells/μL). The linearity of the XN-hsA mode was established up to 938 cells/μL. The cell number obtained using the XN-hsA mode correlated highly with the corresponding microscopic examination. Good correlation was also observed between the DI-60 analyses and manual microscopic classification for all leukocyte types, except monocytes. In conclusion, the combined use of cell counting with the XN-hsA mode and automated morphological analyses using the DI-60 mode is potentially useful for the automated analysis of BF cells.

15 keywords...

hide…