Anterior Ethmoidal Artery, Keros Classification, and Supraorbital Ethmoidal Cells Volume in Paranasal Sinus CT scans: Exploring Relationships and Correlations

Abstract

Objective

This study aims to evaluate the relationships between the anterior ethmoidal artery (AEA) notch landmark, Keros classification, and the presence and volume of supraorbital ethmoidal cells (SOECs) to determine their anatomical correlations and surgical implications.

Materials and Methods

A retrospective analysis of 102 paranasal sinus CT scans (204 sides) was performed. The AEA notch landmark was classified into three degrees based on its vertical distance from the skull base. The presence and volume of SOECs were recorded, and Keros classification was assessed. Correlations among these parameters were analyzed using Spearman's rank correlation coefficient, Chi-square, and Fisher's exact tests.

Results

SOECs were identified in 39.22% of cases, with a mean volume of 0.3535 ± 0.109 cm³ on the right side and 0.3399 ± 0.054 cm³ on the left. A significant correlation was observed between Keros classification and the AEA notch landmark (Spearman's rho = 0.637, p < 0.05). Additionally, SOEC volume showed a weak but significant correlation with the AEA notch landmark (Spearman's rho = 0.161, p < 0.05). However, there was no significant correlation between SOEC volume and Keros classification (p > 0.05).

Conclusion

AEA positioning is significantly influenced by both its distance from the skull base and SOEC volume, though SOEC volume does not affect Keros classification. These insights contribute to a better understanding of endoscopic sinus surgery anatomy and may help minimize surgical risks.

Keywords

anterior ethmoidal artery - cerebrospinal fluid - functional endoscopic sinus surgery - supraorbital ethmoidal cells

Introduction

Functional endoscopic sinus surgery (FESS) is widely recognized as a safe and effective procedure for treating various sinus diseases, with an overall complication rate of approximately 1%. However, severe complications such as cerebrospinal fluid (CSF) leakage and anterior ethmoidal artery (AEA) injury can result in serious outcomes, including intraorbital hemorrhage and even blindness.[1] Studies from Japan have reported that major complications occur in 0.23 to 1% of FESS cases over a 7-year period, with specific risks including CSF leakage (0.09–0.56%), orbital injury (0.07–0.9%), severe postoperative bleeding (0.18–0.76%), and toxic shock syndrome (0.02–0.17%).

Among these complications, intraoperative hemorrhage due to AEA injury is particularly concerning as it can obscure the surgical field, making it difficult to visualize crucial anatomical landmarks and increasing the risk of additional complications.[2] [3] Therefore, understanding the anatomical relationship between the AEA and surrounding structures is critical for minimizing surgical risks. Keros classification is widely used to assess the depth of the olfactory fossa, which is determined by measuring the height of the lateral lamella of the cribriform plate.[2] [4] A higher Keros classification is associated with an increased risk of lateral lamella injury, potentially leading to CSF leakage.[5] [6] [7]

The presence of SOECs has been linked to variations in skull base height and AEA positioning, suggesting a potential correlation between SOEC volume and Keros classification.[7] [8] [9] The height of the skull base appears to be correlated with the location of the AEA. This correlation is observed through the degree of Keros classification.[7] [8] [9] However, previous studies have primarily focused on the presence of SOECs rather than their volumetric impact on AEA location.

This study aims to analyze the relationship between SOEC volume and AEA positioning by evaluating its association with Keros classification. We employed three-dimensional reconstruction techniques to measure and compare SOEC volume and anatomical variations in AEA positioning.[10] By utilizing paranasal sinus CT scans, we investigated the visualization of the AEA notch, the presence and volume of SOECs, and Keros classification to establish their interrelationships. This analysis provides crucial insights for improving preoperative risk assessment in endoscopic sinus surgery.

Materials and Methods

This retrospective study included high-resolution paranasal sinus CT scans from a tertiary medical center. A total of 102 patients (204 sides) were included, comprising 39 males and 63 females. The images were collected from the PACS system between August 2021 and July 2023. The mean age of males was 16.31 ± 29.33 years (range 11 to 76 years), while the mean age of females was 13.51 ± 22.63 years (range 14 to 78 years). The CT images were reviewed by two radiologists—one with 26 years of experience and the other with 5 years—based on a consensus interpretation.

The study examined the correlation between AEA notch landmarks, the presence and volume of SOECs, and Keros classification using Spearman's rank correlation coefficient, descriptive Chi-square, and Fisher's exact tests.

Measurements

1. Landmark AEA notch: They were bilaterally evaluated and classified as absent or present. The AEA notch was classified into three degrees based on its distance from the skull base. The vertical distance from the AEA notch to the adjacent skull base was measured using multiplanar CT reformats in the coronal plane: 1st degree, AEA running within the skull base (<2 mm) ([Fig. 1A]); 2nd degree, AEA running under the skull base and considered prominent (2–4 mm) ([Fig. 1B]); 3rd degree, AEA running freely at a distance from the skull base (>4 mm) ([Fig. 1C]).[11]

2. Presence and volume of SOECs: SOECs were defined as pneumatization of the orbital plate of the frontal bone lateral to the most medial plane of the lamina papyracea.[9] They were classified as absent or present bilaterally. When SOECs were identified, the volume was calculated ([Fig. 2]).

3. Keros classification of the olfactory fossa: The depth of the olfactory fossa was categorized into three types based on the height of the lateral lamella.[12] [13] [14] [15] Type I: 1–3 mm ([Fig. 3A]); type II: 4–7 mm ([Fig. 3B]); type III: 8–16 mm ([Fig. 3C]).

Fig. 1 The classification of the landmark anterior ethmoid artery (AEA) notch- (A) AEA close to the skull base (<2 mm); (B) AEA slightly below the skull base (2–4 mm); (C) AEA hanging farther from the skull base (>4 mm). Arrowheads mark the AEA notch.

Fig. 2 The supraorbital ethmoid cells (SOECs) are shown. When the SOEC was found (arrow), the volume of the SOEC was calculated using a DICOM viewer; the left one was 0.29 cm³ and the right one was 0.17 cm³.

Fig. 3 Three types of the Keros classification of the olfactory fossa. DICOM viewer was used to measure the height of the lateral lamellae. In some samples of the patients studied, (A) the left one was 1.82 mm and the right one was 2.31 mm, (B) the left was 5.93 mm and the right 4.05 mm, and (C) the left was 9.20 mm and the right 7.71 mm.

Results

Among the analyzed scans, the AEA classifications were distributed as follows: Type 1 (40.2% right, 32.4% left), type 2 (33.3% right, 26.5% left), and type 3 (26.5% right, 41.2% left), with no significant difference (p > 0.05, [Table 1]). Overall, the distribution across 204 sides was 1st degree in 36.3%, 2nd degree in 29.9%, and 3rd degree in 33.8%.

Table 1 Identification of landmark AEA notch, Keros, and presence of SOECs

Side		Right (n = 102)		Left (n = 102)		P
		n	%	n	%
AEA	1st degree	41	40.2	33	32.4	p = 0.085a
	2nd degree	34	33.3	27	26.5
	3rd degree	27	26.5	42	41.2
Keros	Type I	6	5.9	8	7.8	p = 0.191a
	Type II	76	74.5	64	62.7
	Type III	20	19.6	30	29.4
SOEC	Absent	59	57.8	65	63.7	p = 0.473[b]
	Present	43	42.2	37	36.3

Abbreviations: AEA, anterior ethmoid artery; SOECs, supraorbital ethmoid cells.

Notes: ^aChi-square.

^b Fisher's exact test.

Meanwhile, Keros classification showed type I in 5.9% (right) and 7.8% (left), type II in 74.5% (right) and 62.7% (left), and type III in 19.6% (right) and 29.4% (left), also with no significant difference (p > 0.05, [Table 2]).

Table 2 Identification volume of SOECs

Side	Right ( n  = 102)			Left ( n  = 102)			P
Volume SOECs	Mean	Median	Std. Dev.	Mean	Median	Std. Dev.
	0.3535	0.02	0.109	0.3399	0.02	0.054	0.3428

Abbreviation: SOECs, supraorbital ethmoid cells.

SOECs were present in 42.2% (right) and 26.3% (left). Fisher's exact test showed no significant correlation between SOEC presence and AEA classification (p > 0.05, [Table 3]). The volume of SOECs was 0.3535 ± 0.109 cm³ (median 0.02 cm³) on the right and 0.3399 ± 0.054 cm³ (median 0.02 cm³) on the left, with no significant difference (p > 0.05, [Table 4]). However, SOEC volume showed a significant correlation with AEA notch landmark (p = 0.0218, [Table 5]), suggesting that larger SOECs may be associated with a more inferiorly positioned AEA.

Table 3 Correlation between the presence of SOECs and Keros classification

SOEC present	Keros classification						Total		Pa = 0.3928 ρ = −0.603 Z for a 95% CI = −0.196 to 0.0779
	Type I		Type II		Type III
	n	%	n	%	n	%	N	%
Absent	6	7.5	57	71.3	17	21.3	80	100
Present	8	6.5	83	66.9	33	26.6	124	100
Total	14	6.9	140	68.6	50	24.5	204	100

Abbreviation: SOECs, supraorbital ethmoid cells.

Notes: ^aSpearman's rank correlation coefficient test.

Description:

P: Significance level.

ρ: Spearman's coefficient of rank correlation.

Z for a 95% CI: 95% confidence interval for rho.

Table 4 Correlation between the presence of SOECs and landmark of AEA notch

SOEC present	Landmark AEA notch						Total		Pa = 0.0887 ρ = −0.119 Z for a 95% CI = −0.253 to 0.0182
	1st degree		2nd degree		3rd degree
	n	%	n	%	n	%	N	%
Absent	32	40	28	35	20	25	80	100
Present	42	33.9	33	26.6	49	39.5	124	100
Total	74	36.3	61	29.9	69	33.8	204	100

Abbreviations: AEA, anterior ethmoid artery; SOECs, supraorbital ethmoid cells.

Notes: ^aSpearman's rank correlation coefficient test.

Description:

P: Significance level.

ρ: Spearman's coefficient of rank correlation.

Z for a 95% CI: 95% confidence interval for rho.

Table 5 Correlation between Keros classification and landmark of AEA notch

Keros classification	Landmark AEA notch						Total		Pa = 0.000 ρ = 0.637 Z for a 95% CI = −0.547 to 0.712
	1st degree		2nd degree		3rd degree
	n	%	n	%	n	%	N	%
Type I	10	71.4	4	28.6	0	0	14	100
Type II	64	45.7	53	37.9	23	16.4	140	100
Type III	0	0	4	8	46	92	50	100
Total	74	36.3	61	29.9	69	33.8	204	100

Abbreviation: AEA, anterior ethmoid artery.

Notes: ^aSpearman's rank correlation coefficient test.

Description:

P: Significance level.

ρ: Spearman's coefficient of rank correlation.

Z for a 95% CI: 95% confidence interval for rho.

The correlation between the presence of SOECs and Keros classification was insignificant (p = 0.3928, [Table 6]), supporting the hypothesis that SOEC pneumatization does not directly affect the lateral lamella of the cribriform plate. In contrast, a significant correlation was found between Keros classification and AEA notch landmark (p < 0.0001, [Table 7]), with a moderate positive correlation coefficient (r = 0.637), suggesting that an increased Keros type is associated with a more inferiorly positioned AEA.

Table 6 Correlation between the volume of SOECs and landmark AEA notch

Sample size	204
Spearman's coefficient of rank correlation (rho)	0.161
Significance level	p = 0.0218
95% confidence interval for rho	0.0237 to 0.292

Abbreviations: AEA, anterior ethmoid artery; SOECs, supraorbital ethmoid cells.

Table 7 Correlation between volume of SOECs and Keros classification

Sample size	204
Spearman's coefficient of rank correlation (rho)	0.105
Significance level	p = 0.1342
95% confidence interval for rho	−0.0326 to 0.239

Abbreviation: SOECs, supraorbital ethmoid cells.

Discussion

One of the complications of the FESS procedure is severe postoperative bleeding, occurring in around 0.8 to 0.76% of cases.[16] This bleeding is commonly due to AEA injury, which is influenced by surrounding anatomical structures such as the orbital lamina of the frontal bone, the lateral lamella of the cribriform plate, and the pneumatization of the SOECs. Notably, SOECs are not consistently present in every individual and may differ between sides. Their presence can affect the positioning of the ethmoidal fovea and cause the AEA to be positioned inferiorly to the cranial base, increasing the risk of injury. Studies have shown that SOEC pneumatization serves as a landmark for AEA location, even in cases of aberrant anatomy. The AEA is typically positioned at the posterior border of the SOEC, even in cases of pathological SOEC expansion.[17]

Identifying the AEA course before surgery is essential to reduce complications during functional endoscopic sinus surgery (FESS). In our study, we classified the AEA notch landmark using Lannoy et al's method.[11] A third-degree notch—defined as AEA located >4 mm below the skull base—was observed in 26.5% of right sides and 41.2% of left sides. Across both sides (204 total), the distribution of AEA notch classification was 1st degree in 36.3%, 2nd degree in 29.9%, and 3rd degree in 33.8%. These findings align with those reported by Taha et al, who found that more than one-third of patients exhibited significant AEA descent associated with supraorbital pneumatization.[3] This highlights a considerable risk of vascular injury in a substantial portion of patients, underscoring the importance of individualized preoperative imaging evaluation.

We examined the correlations between AEA notch classification, Keros classification, and SOEC characteristics. SOECs were identified in 39.2% (80 out of 204 sides), similar to the 44% prevalence reported by Jang et al,[18] while the AEA canal was present in 37.6 to 45.6% of patients, and the AEA sulcus in 53.5 to 61.2%. These findings align with those of Jang et al, who reported a presence of 44% (68 out of 156 sides).[18] Although no significant correlation was found between SOEC presence and AEA notch classification, a statistically significant association emerged between SOEC volume and the AEA notch degree. This suggests that the size—rather than the presence—of SOECs may influence the vertical position of the AEA. Smaller SOECs may have little impact, whereas larger SOECs may push the AEA further from the skull base, increasing its vulnerability during surgery. This finding is consistent with Wang et al,[6] who demonstrated that larger SOEC volumes moderately correlated with greater AEA descent (r = 0.28–0.45, p < 0.001), and with Simmen et al,[13] who observed a descent of up to 3.7 mm in cadaveric specimens with extensive pneumatization.

No significant correlation was found between SOEC presence or volume and Keros classification. Since SOECs originate from the suprabullar recess of the frontal bone and are defined as the superolateral pneumatization of the orbital plate, they primarily influence the space between the skull base and the superior orbital border. These structures expand the space between the cranial plate and the superior orbital border, rather than the lateral lamella of the cribriform plate, which defines Keros classification. Thus, SOECs appear not to influence the depth of the olfactory fossa. Our findings are consistent with the Indian study by Singh et al,[15] which reported that SOEC presence significantly altered AEA location but had no direct influence on olfactory fossa depth.

However, we observed a significant correlation between Keros classification and the AEA notch landmark. A higher Keros classification may lead to a more suspended or hanging AEA course. This may increase the difficulty of endoscopic procedures and raise the risk of vascular or dural injury during skull base surgery. Phayvanh et al (2017) also highlighted that deeper Keros types correspond with greater AEA descent and frontal recess width, reinforcing the importance of integrating these parameters into preoperative planning.[14]

Although this study focused on AEA notch classification and SOEC volume, we did not assess AEA angle or intranasal length due to imaging constraints. Future studies using multiplanar reconstruction or surgical validation could help further characterize these relationships.

These anatomical insights emphasize the importance of evaluating SOEC morphology, AEA trajectory, and Keros classification together in preoperative imaging. A comprehensive understanding of these variations may improve surgical planning and reduce the incidence of complications during endoscopic sinus and skull base procedures.

Conclusion

This study concluded that there is a significant correlation between the classification of the AEA notch and both the volume of SOECs and the degree of Keros classification. However, the presence of SOECs did not show a significant correlation with AEA notch classification. Neither the presence nor the volume of SOECs significantly correlates with the degree of Keros classification. Given the significant correlation between the degree of Keros classification and the AEA notch landmark, surgeons should consider these relationships to minimize the risk of AEA injury and enhance safety during endoscopic sinus surgery.

Conflict of Interest

None declared.

Acknowledgments

The authors would like to thank the Faculty of Medicine, Universitas Airlangga, and Universitas Airlangga Hospital, Surabaya, Indonesia, for all the technical support that has been provided. In addition, the authors would also like to thank the Institute of Research and Community Service of Universitas Airlangga (LPPM) for financial support through Airlangga Research Fund 2023 in the Basic Research Excellence scheme with contract number 2419/UN3. 1.1/PT/2023.

• References

• 1 Wong DKC, Shao A, Campbell R, Douglas R. Anterior ethmoidal artery emerging anterior to bulla ethmoidalis: an abnormal anatomical variation in Waardenburg's syndrome. Allergy Rhinol (Providence) 2014; 5 (03) 168-171
• 2 O'Brien Sr WT, Hamelin S, Weitzel EK. The preoperative sinus CT: avoiding a “CLOSE” call with surgical complications. Radiology 2016; 281 (01) 10-21
• 3 Taha MA, Hall CA, Zylicz HE. et al. Variability of the anterior ethmoid artery in endoscopic sinus surgery. Ear Nose Throat J 2022; 101 (04) 268-273
• 4 Chan Y, Goddard JC. Endoscopic sinus surgery. In: K.J. Lee's Essential Otolaryngology: Head and Neck Surgery, 12e. McGraw-Hill Education; 2019. Accessed at: accesssurgery.mhmedical.com/content.aspx?aid=1172370691
• 5 Özdemir A, Bayar Muluk N. The important adjacent structures for anterior ethmoidal artery in FESS: Anterior ethmoidal artery canal angle, supraorbital ethmoid cells and Keros classification. J Clin Neurosci 2022; 98 (98) 207-212
• 6 Wang H, Li A, Bie T. et al. Correlation analysis of the pneumatization of the supraorbital ethmoid cell and the position of the anterior ethmoidal artery. Ear Nose Throat J 2023; 0 (321) 1-7
• 7 Poteet PS, Cox MD, Wang RA, Fitzgerald RT, Kanaan A. Analysis of the relationship between the location of the anterior ethmoid artery and Keros classification. Otolaryngol Head Neck Surg 2017; 157 (02) 320-324
• 8 Sağlam T, Deniz MA, Turmak M, Hattapoğlu S, Akbudak İ, Tekinhatun M. Relation between anterior ethmoidal artery course on computed tomography and supraorbital ethmoid cell and Keros classification. Eur Arch Otorhinolaryngol 2024; 281 (03) 1293-1299
• 9 Feng Y, Yan S, Wang J, Liu Y, Li X. [The role of supraorbital ethmoid cell and the significance of a proposed classification of the anterior ethmoid artery in endoscopic sinus surgery] [Chinese]. Lin Chuang Er Bi Yan Hou Tou Jing Wai Ke Za Zhi 2021; 35 (06) 495-500
• 10 Hacl A, Costa ALF, Oliveira JM. et al. Three-dimensional volumetric analysis of frontal sinus using medical software. J Forensic Radiol Imaging 2017; 11: 1-5
• 11 Lannoy-Penisson L, Schultz P, Riehm S, Atallah I, Veillon F, Debry C. The anterior ethmoidal artery: radio-anatomical comparison and its application in endonasal surgery. Acta Otolaryngol 2007; 127 (06) 618-622
• 12 Cunha BSA, Garcia DM, Cruz AAV. The relation of the anterior ethmoidal foramen to the cranial base. Ophthalmic Plast Reconstr Surg 2023; 39 (06) 617-620
• 13 Simmen D, Raghavan U, Briner HR, Manestar M, Groscurth P, Jones NS. The surgeon's view of the anterior ethmoidal artery: a cadaveric and computed tomographic study. Rhinology 2006; 44 (02) 164-168
• 14 Sjogren PP, Waghela R, Ashby S, Wiggins RH, Orlandi RR, Alt JA. International Frontal Sinus Anatomy Classification and anatomic predictors of low-lying anterior ethmoidal arteries. Am J Rhinol Allergy 2017; 31 (03) 174-176
• 15 Singh P, Srivastava M, Yadav YR, Tripathi A. Radiological correlation between the anterior ethmoidal artery and the supraorbital ethmoid cell. Indian J Otolaryngol Head Neck Surg 2010; 62 (02) 121-125
• 16 Koizumi M, Suzuki S, Matsui H, Fushimi K, Yamasoba T, Yasunaga H. Trends in complications after functional endoscopic sinus surgery in Japan: a comparison with a previous study (2007-2013vs. 2013-2017). Auris Nasus Larynx 2020; 47 (05) 814-819
• 17 Li M, Sharbel DD, White B, Y. Tadros S, Kountakis SE. Reliability of the supraorbital ethmoid cell vs Keros classification in predicting the course of the anterior ethmoid artery. Int Forum Allergy Rhinol 2019; 9 (07) 821-824
• 18 Jang DW, Lachanas VA, White LC, Kountakis SE. Supraorbital ethmoid cell: a consistent landmark for endoscopic identification of the anterior ethmoidal artery. Otolaryngol Head Neck Surg 2014; 151 (06) 1073-1077

Publication History

Article published online:
10 February 2026