Adelsperger Amelia R, Phillips Evan H, Ibriga Hilda S, Craig Bruce A, Green Linden A, Murphy Michael P, Goergen Craig J
Weldon School of Biomedical Engineering, Purdue University, West Lafayette, Indiana.
Department of Statistics, Purdue University, West Lafayette, Indiana.
Physiol Rep. 2018 Apr;6(8):e13668. doi: 10.14814/phy2.13668.
Abdominal aortic aneurysms are pathological dilations that can suddenly rupture, causing more than 15,000 deaths in the U.S. annually. Current treatment focuses on observation until an aneurysm's size warrants surgical intervention. Thus, there is a need for therapeutic intervention to inhibit growth of smaller aneurysms. An experimental aneurysm model that infuses angiotensin II into apolipoprotein E-deficient mice is widely used to investigate underlying pathological mechanisms and potential therapeutics, but this model has two caveats: (1) aneurysms do not always form, and (2) aneurysm severity and growth is inconsistent among animals. Here we use high-frequency ultrasound to collect data from angiotensin II-induced aneurysms to develop prediction models of both aneurysm formation and growth. Baseline measurements of aortic diameter, volume/length, and strain were used with animal mass and age in a quadratic discriminant analysis and logistic regression to build two statistical models to predict disease status. Longitudinal ultrasound data were also acquired from mice with aneurysms to quantify aneurysm diameter, circumferential strain, blood flow velocity, aneurysm volume/length, and thrombus and open-false lumen volumes over 28 days. Measurements taken at aneurysm diagnosis were used with branching artery information to produce a multiple linear regression model to predict final aneurysm volume/length. All three statistical models could be useful in future aneurysm therapeutic studies to better delineate the effects of preventative and suppressive treatments from normal variations in the angiotensin II aneurysm model.
腹主动脉瘤是一种病理性扩张,可能会突然破裂,在美国每年导致超过15000人死亡。目前的治疗方法侧重于观察,直到动脉瘤大小需要进行手术干预。因此,需要进行治疗干预以抑制较小动脉瘤的生长。一种将血管紧张素II注入载脂蛋白E缺陷小鼠的实验性动脉瘤模型被广泛用于研究潜在的病理机制和潜在的治疗方法,但该模型有两个缺陷:(1)动脉瘤并不总是形成,(2)动物之间动脉瘤的严重程度和生长情况不一致。在这里,我们使用高频超声从血管紧张素II诱导的动脉瘤中收集数据,以建立动脉瘤形成和生长的预测模型。在二次判别分析和逻辑回归中,将主动脉直径、体积/长度和应变的基线测量值与动物体重和年龄一起使用,以建立两个统计模型来预测疾病状态。还从小鼠动脉瘤中获取纵向超声数据,以量化28天内动脉瘤直径、圆周应变、血流速度、动脉瘤体积/长度以及血栓和真假腔体积。在动脉瘤诊断时进行的测量与分支动脉信息一起用于生成多元线性回归模型,以预测最终动脉瘤体积/长度。所有这三个统计模型在未来的动脉瘤治疗研究中可能会有用,以便更好地从血管紧张素II动脉瘤模型的正常变化中区分预防和抑制治疗的效果。