Skip to main content
Fig. 9 | BMC Veterinary Research

Fig. 9

From: Comparison of Sysmex XN-V body fluid mode and deep-learning-based quantification with manual techniques for total nucleated cell count and differential count for equine bronchoalveolar lavage samples

Fig. 9

Agreement between manually and via artificial intelligence (AI) algorithm counted neutrophils on images of virtual cytospin preparation slides when performing a 200-cell differential count per sample. The graph on the left (A) is a Passing-Bablok regression analysis with intercept 2.2 (0.63 to 3.64)* and slope 1.0 (0.89 to 1.08)*. The graph on the right (B) is a Bland–Altman difference plot. The thin horizontal grey line (0 at the y-axis) is the line of identity, and the thick black line indicates the bias (mean difference between methods), with its confidence intervals as thin blue dashed lines. The black dashed horizontal lines are the 95% limits of agreement with their 95% confidence intervals as the thin blue dashed lines. The mean difference is 1.6 (-0.56 to 3.86)* cells, the Lower Limit of Agreement is -16.3 (-20.06 to -12.47)* cells, the Upper Limit of Agreement is 19.6 (15.76 to 23.36)* cells. *Numbers in parentheses are 95% confidence intervals

Back to article page