Fig 1.
Measurement of RBC, fibrin/platelet, and WBC composition using the Positive Pixel Count (PPC) algorithm of the ImageScope program in an RBC-rich clot (A), a mixed clot (B), and a fibrin/platelet-rich clot (C). To measure the area of RBC and fibrin/platelet in a digitized clot image, hue value (0.05), hue width (0.4), saturation (0.04), the intensity of strong positive pixels (Isp, 100), and the intensity of weak positive pixels (Iwp, 215) were set to the fixed values. The intensity of positive pixels (Ip) was only adjusted in a best image of the 10 clot images differently digitized from 135 to 180 of Ip values to obtain the clearest discrimination of RBC and fibrin/platelet in the individual clot images.
Table 1.
Patient characteristics.
Table 2.
Number of trials for clot retrieval and successful recanalization.
Fig 2.
Comparison of clot composition (red blood cells [RBCs], fibrin/platelets, white blood cells [WBCs]) in total retrieved clots (A) and in total included patients according to stroke subtype. LAA: Large artery atherosclerosis.
Table 3.
Frequencies of clot types by the stroke subtypes in all retrieved clots.
Fig 3.
Comparison of clot composition (red blood cells [RBCs], fibrin/platelets, white blood cells [WBCs]) according to recanalization status (A), the presence of clot signs (hyperdense MCA sign on initial computed tomography and/or blooming artifact on initial gradient-echo magnetic resonance images) (B), and the retrieval devices (C).