For this study, archived images of patients undergoing thoracic MDCT examination for various reasons at Gaziantep University Faculty of Medicine between 2017 and 2021 were reviewed retrospectively. Among these, 313 images that best suited our purpose were chosen for the study.
Exclusion criteria: Poor image quality due to motion artifacts or insufficient distribution of contrast material, absence of aortic arch and its branches in the field of view and patients with supra-aortic branches coming off from a single trunk which could unfavorably affect angle measurements were excluded.
Of 313 images, 142 images were from females and 171 from males. The mean age of the patients was 55.77±19.48 years. All measurements were obtained by the same investigator. The data were anonymized to avoid identification of the patients.
A 64-detector MDCT (Light Speed VCT XTe; General Electric, Milwaukee, USA) was used to acquire patient images. 120 ml of non-ionic contrast agent with an iodine concentration of 300 mgI/ml was injected into the right or left antecubital vein with the help of an automated injector (Covidien LF OptiVantage DH, Ohio, USA) as a bolus at a rate of 4 ml/sec, followed by 40 ml of saline at a rate of 4 ml/sec. The following parameters were used for all scans: collimation, 40 mm (64x0.625); pitch value, 1; tube rotation time, 0.35 sec; X-ray tube operating at 100-120 kV and 150-600 milliampere; detector thickness, 0.625 mm; reconstruction interval, 0.625 mm.
Three-dimensional (3D) reconstructions of two-dimensional (2D) MDCT images were made using the 3D Slicer, version 4.13.0 (https://www.slicer.org), an open source software package. Following transfer of the images to the 3D Slicer software, automatic thresholding was performed but since this process was insufficient, bone and other anatomical structures around the aorta were removed manually on the software. All images were viewed in the oblique coronal plane, where the angulation of the aortic arch and its branches can be best seen. The oblique coronal plane was determined using the guiding lines in the 3D reconstruction of the 3D slicer software.
First, transverse lines were drawn tangential to the outer and inner curvature of the aortic arch in order to classify aortic arch types [20]. The location of the brachiocephalic trunk was determined using these lines and the aortic arch configuration was divided into three types [7,10,16,17]. Then, these configurations were divided into subtypes based on the location of other supra-aortic vessels (Fig. 1):
Aortic arch configuration
Type 1: Brachiocephalic trunk (BCT) is above the upper line.
Type 1a (normal configuration): if all supra-aortic branches (left subclavian artery (LSA), left common carotid artery (LCCA) and BCT) are above the upper line,
Type 1b: if LSA is between the two lines and others are above the upper line,
Type 1c: if BCT is above the upper line and other are between the two lines.
Type 2: Brachiocephalic trunk is between the two lines.
Type 2a: if LSA is above the upper line and other branches are between the two lines,
Type 2b: if BCT is between the two lines and others are above the upper line,
Type 2c: if all supra-aortic branches are between the two lines,
Type 2d: if common carotid artery (CCA) is above the upper line and other branches are between the two lines.
Type 3: Brachiocephalic trunk is below the lower line.
Type 3a: if LSA is between the two lines and others are below the lower line,
Type 3b: if BCT is below the lower line and others are between the two lines.
The midpoints of both the ascending and descending aorta (at the point where the truncus pulmonalis emerges) were determined manually. The aortic arch angle was measured by drawing lines connecting these midpoints to the top of the arch. In the same way, the midpoints of the supra-aortic branches were marked manually, and their angles were measured where they originate from the aortic arch, parallel to the ground (Fig. 2). Since different methods were used for angle measurements, we could not identify a standard method used for this purpose [2,4,15,18,23,26]. Therefore, we also chose to measure the angles of aortic arch and its branches parallel to the ground in order to avoid measurement errors in the face of variations.
Statistical analysis
Descriptive statistics of the study data were summarized as mean and standard deviation for numerical variables, and frequency and percentage for categorical variables. Whether the angle measurements followed a normal distribution was examined using Shapiro-Wilk test. Except for aortic arch angle (AAA), all variables were normally distributed (p<0.05). For categorical variables, Independent samples t-test and Mann-Whitney U test were used for comparisons between two groups, and Analysis of Variance (ANOVA) and Kruskal-Wallis test to compare more than two groups. Differences among the groups identified by ANOVA and Kruskal-Wallis test were further analyzed using Tukey and Dunn tests. The results of statistical analyses were reported as mean ± standard deviation for the variables following a normal distribution and as median (Q1-Q3) for non-normally distributed variables. In addition, Chi-square test was used to examine the differences among the groups in terms of categorical variables. The associations between numerical variables were investigated using Pearson correlation and Spearman correlation analyses. All statistical analyses were performed using SPSS for Windows, version 22.0 (IBM Corp., Armonk, NY). The statistical significance level was set at p<0.05.