Anatomical characteristics and morphometric analyses of the internal carotid artery using retrospective angiographic images

Anatomical variations of the internal carotid artery (ICA) can cause complications during endovascular treatments. Therefore, the aim of this study was to investigate the features of the ICAs obtained from 2D digital subtraction angiography (2D DSA). The morphometrics and angles of the ICA segments from 2D DSA images from a total of 100 patients (45 males and 55 females) were investigated. The lengths (L1-3), angles (A1-3), and diameters (D1-6) through the ICA measurement points (five segments: C2 [petrous], C3 [lacerum], C4 [cavernous], C5 [clinoid], C6 [ophthalmic]) were systemically recorded by two observers. All measured parameters were compared for both sexes and sides. The lengths (mm) and angles of the ICA were 7.20 ± 2.22 (L1), 15.71 ± 2.32 (L2), 10.99 ± 1.66 (L3) and 109.31 ± 17.77 (A1), 107.87 ± 20.51 (A2), 80.81 ± 16.33 (A3), respectively. There were no differences in the angulations of the A1–A3 segments between the sexes (p > 0.05). The L2 (C3–C4) was significantly longer in males, but only the left side of females showed a greater length (p < 0.05). The averaged mean diameters of the ICA (D1-6) in both sexes were 4.17 ± 0.55 mm. However, the diameters of the D1–D6 from left and right sides of males except the right D6 were significantly longer than those of females (p < 0.05). Also of significance was that the right diameters of females (D1) and males (D1 and D2) were shorter compared to left sides. This study demonstrates some differences in lengths, diameters, and angles in both sexes and sides of the normal ICA. Neurosurgeons and neurointerventionalists should be aware of such variations when operating, performing interventional procedures, and interpreting 2D DSA.


Introduction
The internal carotid artery (ICA) is the main blood supply to the anterior and middle parts of the brain. It can be divided into seven segments based on the classification reported by Bouthillier coworkers [3]. These segments consist of a C1 (cervical), C2 (petrous or horizontal), C3 (lacerum), C4 (cavernous), C5 (clinoid), C6 (ophthalmic or supraclinoid), and C7 (communicating or terminal) parts. Many studies have reported that anatomical variations of the ICA can be associated with various neurovascular diseases and thus, can consequently complicate endovascular and surgical approaches [4,10,12,14,15,17,20]. For example, variant ICAs have been reported to result in complications during treatment of traumatic cavernous sinus fistulas (TCCF) [5,13,21]. Therefore, the anatomical characteristics and 1 3 morphometrics of individual ICAs are important to consider prior to, for example, aneurysm clipping, angioplasty, and stenting. Previous studies have also documented that the morphometrics of the ICA for cadaver specimens and radiological images are also different among various populations [1,8,12,19].
It is known that the cerebral angiography with twodimension (2D) and three-dimension (3D) imaging is the gold standard for the diagnosis of intracranial aneurysms and other neurovascular diseases [7]. It can be performed in both 2D and 3D imaging of all four vessels that supply the blood to the brain. Differences or variations of the ICA using 2D digital subtraction angiography data (2D DSA) have been reported between the sexes, sides, and ages of the patients [14,16,18]. Therefore, this study aimed to investigate the ICA variations and morphometry from 2D DSA images in a Thai population.

Ethical approval and patient selection
This retrospective study on the characteristics and morphometrics of ICA was approved by the Khon Kaen University Ethics Committee for Human Research (code: HE641277). ICA morphometric analyses from 2D images were performed from January 2019 to December 2020. All parameters were performed and evaluated within a picture archiving and communication system (PACs) program. One hundred patients were studied with the following inclusion criteria: subarachnoid hemorrhage patients who underwent cerebral angiography for intracranial aneurysm diagnosis and had both ICA 2D images without any visible pathology within the ICA. The length of the vertical part of the ICA-CS was measured on 3D images which were produced from the raw data recorded in the DSA unit (Siemens Artizee, Germany), Srinagarind Hospital, Khon Kaen University, Thailand. The patients' information was retrieved from the interventional neuroradiology unit, collected from January 2020 to May 2021. Fifty-two patients who had undergone cerebral angiography were included in this study. All patients had ICA 3D images from both the right and left sides. Patients with any pathology of the skull base or who did not have complete ICA 2D imaging were excluded from the study. The other exclusion criteria were (1) not clearly determine all points of measurements such as particularly C5 and C6 that are very closed to each other, (2) overlapping bending of ICA especially C2 in some cases which is difficult to determine the point for measurements of diameter and length, and (3) patients were diagnosed as atherosclerosis. We have already put more information in the revised MS.

Morphometric measurements
The measurements were performed as follows; (1) selected the 2D ICA image with optimal contrast media filling in vessels from a lateral view, (2) magnified images as much as possible to view all points (as shown in Fig. 1), (3) marked the locations of C2-6 and (4) measured the length of L1-L3 (Fig. 1B), A1-A3 (Fig. 1C), and D1-D6 (Fig. 1D), respectively. All measurements were performed using PACS image tools. The elliptical region of interest was used to localize the C2-C6 while the lengths of L1-L3 and D1-D6 were performed using the ruler tool. The measurement of A1-A3 was performed as previously reported [17]. The angle measurement tool was used to analyze the angles between the different parts of the ICA. Then the angles of each bend were measured and recorded in triplicate. The 21 cases were used to determine the reliability between intra-and inter-observers.

Statistical analysis
Statistical analyses were performed using IBM SPSS Statistics 21 and the data were expressed as mean ± SD. The Shapiro-Wilk test was used to determine the data distribution. Independent sample t test and Mann-Whitney U test were used to analyze the differences among all parameters of normal and abnormal distribution data sets. An intraclass correlation coefficient (ICC) was used to analyze the reliability between observers. In addition, the analysis of age and sex correlations was performed using Spearman's rho. A p value < 0.05 was considered significant.

Demographic data of population
One hundred patients (forty-five males and fifty-five females) were recruited to determine the morphometrics 1 3 of the ICA from 2D DSA images. Their mean age was 53 ± 13.28 years old. The age of the female patients ranged from 44 to 81 years and the age of the males ranged from 15 to 77 years old.

Reliability between observers
The reliability between observers showed that there was high reliability in both intra-(0.97-1.00) and inter-observers (0.95-1.00). There was no significant difference between them. This high reliability had coefficients of 0.92-0.99 and 0.97-0.99.

ICA morphometry of females and males for various age ranges
Tables 1, 2, and 3 show the morphometrics of the ICA including the lengths, angles, and diameters between females and males from different age ranges (interval ages, 10 years). The age range of patients in this study was 15-85 years old. No female patient was less than 35 years old. It was found that the maximal mean length of the ICA in both sexes was that of L2. The mean length in males aged ≥ 76 years old was 17.45 ± 1.40 mm and that of females aged 66-75 years old was 15.84 ± 2.08 mm. The shortest length of the ICA was for L1 in both sexes with an age range of 36-45 years old and the values for this part in males and females were 5.78 ± 0.65 mm and 5.96 ± 1.16 mm, respectively.
The angle measurements for the various segments of the ICA are shown in The results of the diameter of the ICA for its various segments are shown in Table 3. These demonstrate that the diameter of D4, D5, and D6 in females gradually increased with age. The smallest diameter was found in the age range of 36-45 years old and the largest diameter was found in the age range of 76-85 years old. In males, the diameter of D1 Fig. 1 The lateral view of 2D ICA images, A the measurement points from C2-C6, B the measurement length, C the measurement angle, and D the measurement diameter, respectively 1 3 slightly decreased with age. The largest diameter in males was observed in the age range of 15-35 years old while the age range of ≥ 76 years old was found to have the smallest diameters.

The difference of length, angle, and diameter of ICA between sexes
The comparisons of the lengths and angles of the ICAs between females and males are shown in Table 4. It was found that the L2 of males on both sides was significantly longer than that of females (p < 0.05). The left L2 in males (16.95 ± 2.36 mm) was longer than that of females (16.021 ± 2.23 mm) (p = 0.047) while right L2 of male (15.58 ± 2.13 mm) was also longer than that of female (14.51 ± 1.92 mm) significantly (p = 0.01). However, there was no significant differences in females and males for both sides (p > 0.05). The comparison of diameter of the various segments of ICA (D1-D6) between males and females is shown in Table 5. It was found that the diameter of D1-D6 on left sides and D1-D5 on the right sides of males was significantly larger than that of females (p < 0.05).

The difference of length, angle, and diameters of left and right ICAs
The lengths (L1-L3) and angles (A1-A3) between left and right ICAs of males and females are shown in Table 6. The results showed that the length of the left L2 ICA (16.95 ± 2.36 mm) in males was significantly different from the right ICA (15.58 ± 2.13 mm) (p = 0.005). In contrast, the left L2 ICA in females (16.02 ± 2.23 mm) was also significantly different (p < 0.001) from the right ICA (14.51 ± 1.92 mm). The other parameters showed no significant differences. In addition to length, the angle of the left A1 in males (109.14 ± 17.56 degrees) showed a significant difference (p = 0.005) compared to the right A1 (106.72 ± 16.33 degrees). Moreover, the angle of the left A1 of females showed a similar trend as males with a significant difference (p = 0.002). The comparison of the ICA diameter between left and right sides of males and females is demonstrated in Table 6. It was found that the diameter of the left ICA at D5 (4.07 ± 0.67 mm) in females was significantly different (p = 0.039) from right sides (3.80 ± 0.68 mm). In males, the left ICA at D1 (4.70 ± 0.51 mm) and D2

Correlation of sex and age on the morphometrics of ICA
The correlation analysis revealed that sex had a correlation with D1-D5 (p < 0

Discussion
It has been demonstrated that the morphometrics of the ICA are different among various in many populations [1,8,12,19]. The data source for ICA analysis used in this study was from 2D DSA. Our study differed from previous studies in that we have investigated differences between the sexes. The mean angles of A1 in Thai males (106.72 degrees) and females (102.10 degrees) were more similar than that of an Indian population (102 degrees) [19].  (140 ICAs) from computed tomography angiography (CTA) had differences in sides and sex [1]. In comparison, the diameters of the ICA (D1, D3, D4, D5, and D6 in this study) were comparable and could be compared with C2, C3, C4, C5, and C6, respectively, as described in a previous study [1]. Most of the ICA diameters (C2-C5) in the Romanian patients were larger than that of Thais (D1-D6) although C6 was smaller than (D6) as observed in this study. This is the first study to measure the lengths of the ICA (L1, L2, and L3) in Thais according to previous classifications [3].
The results of this study demonstrated that the diameter of ICA in males was larger than that in females and that the length of the L2 or the length of C3 to C4 in males was also longer than in females. Previous studies have also found that sex differences affected the diameter and the vertical lengths of the ICA [11,14,18]. Hypertension was reported to associate with a smaller ICA diameter. Moreover, it has been shown that factors such as age, weight, and height are associated with the diameter of the femoral artery [18]. Additionally, a patient's height and skull structure have been associated with the size and length of the ICA [1]. However, we did not evaluate these factors because it was not the main objective of our study. It has been shown that age and hemodynamics are both important factors for the variation of the basilar artery. However, this study observed less correlation between age and the morphometric parameters of the ICA. The small sample size and the combination of both sexes in the analysis might also produce some effects. In addition, the increased diameter of the D4-D6 in female patients with increasing age might be due to the reduction of intraluminal elastin [9]. It has also been reported that cardiac hemodynamics and blood flow in the left and right carotid arteries are different with the flow being higher on left sides [2,6]. These factors might also contribute to the significant differences of diameters, lengths, and angles between the left and right ICAs as seen in this study.
In conclusion, the morphometry of the ICA in our population was found to be slightly different from other studied ethnicities. The comparison of morphometric data between sex and side of the ICA was also elucidated.