Canine Cerebrospinal Fluid Analysis Using Two New Automated Techniques: The Sysmex XN-V Body Fluid Mode and an...

Area of Science:

• Veterinary Diagnostics
• Clinical Pathology
• Canine Neurology

Background:

• Cerebrospinal fluid (CSF) analysis is crucial for diagnosing neurological conditions in dogs.
• Current manual methods for total nucleated cell count and differential cell counts are time-consuming.
• There is a need for faster, automated diagnostic methods in veterinary medicine.

Purpose of the Study:

• To evaluate the accuracy of automated methods for canine CSF analysis.
• To compare the Sysmex XN-V body fluid mode and a deep-learning algorithm with manual cell counting.
• To assess the feasibility of using automated systems for routine CSF diagnostics in dogs.

Main Methods:

• Comparison of manual cell counting with automated methods (Sysmex XN-V, Olympus VS200 deep-learning algorithm).
• Analysis of 161 canine CSF samples for total nucleated cell count.
• Analysis of 65 CSF samples with pleocytosis for differential cell counts.
• Manual re-gating of Sysmex XN-V data due to initial incorrect gating.

Main Results:

• The Sysmex XN-V body fluid mode, after manual gate adjustments, showed a mean bias of 15.19 cells/μL for total nucleated cell count.
• Mean biases for the Sysmex XN-V two-part differential count were 4.95% and -4.95%.
• The deep-learning algorithm demonstrated mean biases of -7.25% (lymphocytes), -0.03% (neutrophils), and 7.27% (monocytoid cells).

Conclusions:

• The automated Sysmex XN-V body fluid mode is recommended for total nucleated cell counts in canine CSF after adjusting manufacturer settings.
• Both the Sysmex XN-V two-part differential count and the deep-learning algorithm require further optimization for accurate differential cell analysis in canine CSF.