Semiautomated Segmentation and Measurement of Cytoplasmic Vacuoles in a Neutrophil With General-Purpose Image Analysis Software.

Background: Morphological observation of blood or marrow film is still described nonquantitatively. We developed a semiautomatic method for segmenting vacuoles from the cytoplasm using Photoshop (PS) and Image-J (IJ), called PS-IJ, and measured the relative entire cell area (rECA) and relative areas of vacuoles (rAV) in the cytoplasm of neutrophil with PS-IJ.

Methods: Whole-blood samples were stored at 4°C with ethylenediaminetetraacetate and in two different preserving manners (P1 and P2). Color-tone intensity levels of neutrophil images were semiautomatically compensated using PS, and then vacuole portions were automatically segmented by IJ. The rAV and rECA were measured by counting pixels by IJ. For evaluating the accuracy in segmentations of vacuoles with PS-IJ, the rAV/rECA ratios calculated with results from PS-IJ were compared with those calculated with human eye and IJ (HE-IJ).

Results: The rECA and rAV/ in P1 significantly (P < 0.05, P < 0.05) were enlarged and increased, but did not significantly (P = 0.46, P = 0.21) change in P2. The rAV/rECA ratios by PS-IJ were significantly correlated (r = 0.90, P < 0.01) with those by HE-IJ.

Conclusion: PS-IJ method can successfully segment vacuoles and measure the rAV and rECA, becoming a useful tool for quantitative description of morphological observation of blood and marrow film.