BACKGROUND AND AIM: Microvascular architecture is a variable characterizing early gastric cancer (EGC) against the background. The aims of the present study were to measure morphological variables of the microvessels and to compare the variables between EGC and the background. METHODS: Narrow band imaging (NBI)-equipped magnifying endoscopic pictures from 32 patients with EGC were used. The endoscopic pictures were taken under maximal magnification and processed for the microvessels in an in-focus area after correction of image distortion. The segmented microvessels were numbered for microvessel density (counts/mm(2)) and vascular bed area (% ratio of vascular bed against the region of interest). The microvessels were further processed for a set of skeletonized pixels to count the characteristic points, including end-points, crossing points, branching points and connecting points. RESULTS: Microvessels in cancer were found to have a significantly larger connected point number (20.5 ± 6.1, P = 0.0002) than those in the background (17.4 ± 3.9). Numbers of the end-points and branching points were found to be significantly larger in cancer than in the background (end-points 3.6 ± 0.7 for cancer vs 3.3 ± 0.4 for background, P = 0.0005; branching points 0.8 ± 0.4 for cancer vs 0.7 ± 0.2 for background, P = 0.0014). However, microvessel density, vascular bed area and mean diameter did not significantly differ between cancer and the background. CONCLUSION: This finding can be considered to reflect the reported observation of an irregular vascular pattern in gastric cancer. This method may provide a means for microvessel morphometry, regardless of the organ studied.
BACKGROUND AND AIM: Microvascular architecture is a variable characterizing early gastric cancer (EGC) against the background. The aims of the present study were to measure morphological variables of the microvessels and to compare the variables between EGC and the background. METHODS: Narrow band imaging (NBI)-equipped magnifying endoscopic pictures from 32 patients with EGC were used. The endoscopic pictures were taken under maximal magnification and processed for the microvessels in an in-focus area after correction of image distortion. The segmented microvessels were numbered for microvessel density (counts/mm(2)) and vascular bed area (% ratio of vascular bed against the region of interest). The microvessels were further processed for a set of skeletonized pixels to count the characteristic points, including end-points, crossing points, branching points and connecting points. RESULTS: Microvessels in cancer were found to have a significantly larger connected point number (20.5 ± 6.1, P = 0.0002) than those in the background (17.4 ± 3.9). Numbers of the end-points and branching points were found to be significantly larger in cancer than in the background (end-points 3.6 ± 0.7 for cancer vs 3.3 ± 0.4 for background, P = 0.0005; branching points 0.8 ± 0.4 for cancer vs 0.7 ± 0.2 for background, P = 0.0014). However, microvessel density, vascular bed area and mean diameter did not significantly differ between cancer and the background. CONCLUSION: This finding can be considered to reflect the reported observation of an irregular vascular pattern in gastric cancer. This method may provide a means for microvessel morphometry, regardless of the organ studied.
Authors: Jae Pil Han; Su Jin Hong; Hee Kyung Kim; Yun Nah Lee; Tae Hee Lee; Bong Min Ko; Joo Young Cho Journal: Surg Endosc Date: 2015-04-01 Impact factor: 4.584