Medical Section, Hainan Tunchang People's Hospital, Tunchang, 571600 Hainan, China.
Comput Math Methods Med. 2022 Aug 8;2022:1261259. doi: 10.1155/2022/1261259. eCollection 2022.
This research was aimed at exploring the application value of optical coherence tomography (OCT) images under adaptive segmentation algorithm in the early diagnosis of coronary heart disease (CHD). Eighty-two patients with CHD were included, who were to undergo coronary angiography (CAG) to confirm their condition. According to the diagnostic criteria of CHD in the American Coronary Artery Surgery Study (CASS), the patients were divided into the stable plaque group (41 cases) and unstable plaque group (41 cases). Besides, 20 healthy volunteers were selected as the control group, and all of them underwent OCT scans. On the basis of a fourth-order partial differential equation (PDE) and active contour (AC) model, a novel adaptive image segmentation algorithm PDE-AC was constructed and used for OCT image processing. No significant difference was found in general clinical data and serological indicators in the control group compared to the other two groups ( > 0.05). The lipid plaque length, degree of stenosis, and lipid pool angle, macrophages and intimal erosion, and plaque fissure in the unstable plaque group were highly greater than those in the stable plaque group. The fibrous cap thickness (FCT) was significantly thinner than that in the stable plaque group ( < 0.05). The diagnostic sensitivity, specificity, and accuracy of OCT under PDE-AC algorithm for CHD (91.53%, 84.08%, and 95.38%) were markedly higher than those of single OCT (83.46%, 75.11%, and 88.02%) ( < 0.05). In summary, OCT images under PDE-AC algorithm did better than simple OCT images in the diagnosis of CHD. Lipid plaque length, degree of stenosis, and lipid pool angle, macrophage and intimal erosion, plaque fissure, and FCT were important indicators for judging plaque stability, having the better clinical application value.
本研究旨在探讨自适应分割算法下光学相干断层扫描（OCT）图像在冠心病（CHD）早期诊断中的应用价值。纳入 82 例 CHD 患者，均行冠状动脉造影（CAG）以明确病情。根据美国冠状动脉旁路移植术研究（CASS）CHD 诊断标准，将患者分为稳定斑块组（41 例）和不稳定斑块组（41 例）。另选取 20 例健康志愿者作为对照组，均行 OCT 扫描。在四阶偏微分方程（PDE）和主动轮廓（AC）模型的基础上，构建了一种新的自适应图像分割算法 PDE-AC，并用于 OCT 图像处理。与另外两组相比，对照组的一般临床资料和血清学指标无显著差异（>0.05）。不稳定斑块组的脂质斑块长度、狭窄程度、脂质池角度、巨噬细胞和内膜侵蚀、斑块裂隙高度大于稳定斑块组，纤维帽厚度（FCT）明显小于稳定斑块组（<0.05）。PDE-AC 算法下 OCT 对 CHD 的诊断灵敏度、特异度和准确度（91.53%、84.08%和 95.38%）明显高于单纯 OCT（83.46%、75.11%和 88.02%）（<0.05）。综上所述，PDE-AC 算法下的 OCT 图像在 CHD 诊断中优于单纯 OCT 图像。脂质斑块长度、狭窄程度、脂质池角度、巨噬细胞和内膜侵蚀、斑块裂隙以及 FCT 是判断斑块稳定性的重要指标，具有更好的临床应用价值。
