Background The purpose of this study is to find the high-risk morphological features in type B aortic dissection (TBAD) population and to establish an early detection model.

Methods From June 2018 to February 2022, 234 patients came to our hospital because of chest pain. After examination and definite diagnosis, we excluded people with previous cardiovascular surgery history, connective tissue disease, aortic arch variation, valve malformation, and traumatic dissection. Finally, we included 49 patients in the TBAD group and 57 in the control group. The imaging data were retrospectively analyzed by Endosize (Therevna 3.1.40) software. The aortic morphological parameters mainly include diameter, length, direct distance, and tortuosity index. Multivariable logistic regression models were performed and systolic blood pressure (SBP), aortic diameter at the left common carotid artery (D3), and length of ascending aorta (L1) were chosen to build a model. The predictive capacity of the models was evaluated through the receiver operating characteristic (ROC) curve analysis.

Results The diameters in the ascending aorta and aortic arch are larger in the TBAD group (33.9 ± 5.9 vs. 37.8 ± 4.9 mm, p < 0.001; 28.2 ± 3.9 vs. 31.7 ± 3.0 mm, p < 0.001). The ascending aorta was significantly longer in the TBAD group (80.3 ± 11.7 vs. 92.3 ± 10.6 mm, p < 0.001). Besides, the direct distance and tortuosity index of the ascending aorta in the TBAD group increased significantly (69.8 ± 9.0 vs. 78.7 ± 8.8 mm, p < 0.001; 1.15 ± 0.05 vs. 1.17 ± 0.06, p < 0.05). Multivariable models demonstrated that SBP, aortic diameter at the left common carotid artery (D3), and length of ascending aorta (L1) were independent predictors of TBAD occurrence. Based on the ROC analysis, area under the ROC curve of the risk prediction models was 0.831.

Conclusion Morphological characteristic including diameter of total aorta, length of ascending aorta, direct distance of ascending aorta, and tortuosity index of ascending aorta are valuable geometric risk factors. Our model shows a good performance in predicting the incidence of TBAD.

背景 本研究的目的是寻找B型主动脉夹层（TBAD）人群的高危形态学特征，并建立早期检测模型。

方法 2018年6月至2022年2月，因胸痛来我院就诊的患者234人。经检查确诊，排除既往有心血管手术史、结缔组织病、主动脉弓变异、瓣膜畸形、外伤性夹层的患者。最后，我们将 49 名患者纳入 TBAD 组，将 57 名患者纳入对照组。通过 Endosize (Therevna 3.1.40) 软件对成像数据进行回顾性分析。主动脉形态学参数主要包括直径、长度、直距和迂曲指数。进行多变量逻辑回归模型，并选择收缩压（SBP）、左颈总动脉处的主动脉直径（D3）和升主动脉（L1）的长度来建立模型。

结果 TBAD 组升主动脉和主动脉弓的直径较大（33.9 ± 5.9 对 37.8 ± 4.9 mm，p < 0.001；28.2  ± 3.9 对 31.7 ± 3.0 mm，p  < 0.001）。TBAD 组的升主动脉明显更长（80.3 ± 11.7 对 92.3 ± 10.6 mm，p  < 0.001）。此外，TBAD 组升主动脉的直接距离和迂曲指数显着增加（69.8 ± 9.0 对 78.7 ± 8.8 mm，p < 0.001；1.15 ±  0.05 对 1.17 ± 0.06，p < 0.05)。多变量模型表明，SBP、左颈总动脉 (D3) 的主动脉直径和升主动脉 (L1) 的长度是 TBAD 发生的独立预测因子。基于ROC分析，风险预测模型的ROC曲线下面积为0.831。

结论 全主动脉直径、升主动脉长度、升主动脉直距、升主动脉迂曲指数等形态学特征是有价值的几何危险因素。我们的模型在预测 TBAD 的发生率方面表现良好。