Measurement for the Area of Red Blood Cells From Microscopic Images Based on Image Processing Technology and Its Applications in Aplastic Anemia, Megaloblastic Anemia, and Myelodysplastic Syndrome

Background Aplastic anemia (AA), megaloblastic anemia (MA), and myelodysplastic syndrome (MDS) were common anemic diseases. Sometimes it was difficult to distinguish patients with these diseases. Methods In this article, we proposed one measurement method for the area of red blood cells (RBCs) from microscopic images based on image processing technology and analyzed the differences of the area in 25 patients with AA, 64 patients with MA, and 68 patients with MDS. Results The area of RBCs was 44.19 ± 3.88, 42.09 ± 5.35, 52.87 ± 7.68, and 45.75 ± 8.07 μm 2 in normal subjects, patients with AA, MA, and MDS, respectively. The coefficients of variation were 8.78%, 10.05%, 14.53%, and 14.00%, respectively, in these groups. The area of RBCs in patients with MA was significantly higher than normal subjects ( p < 0.001). Compared with patients with AA and MDS, the area of RBCs in patients with MA was also significantly higher ( p < 0.001). The results of correlation analysis between the area of RBCs and mean corpuscular volume (MCV), mean corpuscular hemoglobin (MCH), MCH concentration (MCHC), and red cell distribution width showed no significant correlations ( p > 0.05). The area under the curve (AUC) results of the Receiver Operating Characteristic (ROC) curves of RBCs area were 0.421, 0.580, and 0.850, respectively, in patients with AA ( p = 0.337), MDS ( p = 0.237), and MA ( p < 0.001). Conclusion Identifying the area of RBCs in peripheral blood smears based on the image processing technology could achieve rapid and efficient diagnostic support for patients with MDS and MA, especially for patients with MA and in combination with MCV. However, a larger sample study is needed to find the cutoff area values.

Abstract licence: CC BY

24 more references, whose wording the paper’s licence keeps private.

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.