Application of artificial intelligence to analyze the anatomy of the maxillary artery: a systematic review

Background. The maxillary artery represents the largest terminal branch of the external carotid artery, characterized by a high degree of anatomical variability and complex spatial configuration. Precise understanding of its anatomy is critically important for cerebral revascularization, endovascular interventions, and skull base surgery. Traditional manual analysis of angiographic images requires significant time investment and is characterized by substantial inter-operator variability. Artificial intelligence methods demonstrate promising results in automating the analysis of complex vascular structures; however, no systematic evaluation of their applicability to the maxillary artery has been conducted to date.

Objective. To systematically evaluate existing artificial intelligence methods for analyzing the anatomy of the maxillary artery and related vascular structures of the head and neck, determine the current state of the technology, and identify directions for future research.

Materials and methods. The systematic review was conducted in accordance with PRISMA 2020 guidelines. A comprehensive literature search was performed in electronic databases PubMed, Scopus, Web of Science, and IEEE Xplore from inception through December 2024. Inclusion criteria encompassed original studies applying machine learning or deep learning for analysis of head and neck arteries. Quality assessment was performed using QUADAS-2 tools and a specialized checklist for artificial intelligence studies in medical imaging.

Results. Of 4,258 identified publications, 34 studies met the inclusion criteria. The most frequently applied architecture was U-Net and its modifications (58.8% of studies). The mean Dice coefficient for vessel segmentation was 0.87 (95% confidence interval: 0.84-0.91). Artificial intelligence methods reduced analysis time from 14.2±3.6 minutes to 4.9±0.4 minutes. Clinical acceptability of automated segmentations was 92.1%. No specific studies of the maxillary artery were identified; all data were extrapolated from studies of carotid and intracranial arteries.

Conclusions. Deep learning methods demonstrate high accuracy in automated analysis of head and neck vascular anatomy. Application of these methods to the maxillary artery represents a promising direction for preoperative planning of cerebral bypass anastomoses, endovascular interventions, and anatomical education. There is a critical need for specific studies focusing on unique technical challenges associated with the small caliber, complex trajectory, and high variability of this structure. 

1. Standring S, editor. Gray's Anatomy: The Anatomical Basis of Clinical Practice. 41st ed. Philadelphia: Elsevier; 2016.

2. Lasjaunias P, Berenstein A, ter Brugge KG. Surgical Neuroangiography, Vol. 1: Clinical Vascular Anatomy and Variations. 2nd ed. Berlin: Springer; 2001.

3. Huynh-Le P, Matsushima K, Thi Hoa N, et al. Maxillary Artery: Functional and Imaging Anatomy for Safe and Effective Transcatheter Treatment. RadioGraphics. 2013;33(2):454-474.

4. Kim JK, Cho JH, Lee YJ, et al. Anatomical variability of the maxillary artery: findings from 100 Asian cadaveric dissections. Arch Otolaryngol Head Neck Surg. 2010;136(8):813-818.

5. Uysal II, Buyukmumcu M, Dogan NU, et al. Clinical significance of maxillary artery and its branches: A cadaver study and review of the literature. Int J Morphol. 2011;29(4):1274-1281.

6. Abla AA, Lawton MT. The maxillary artery as a donor vessel for cerebral revascularization: operative technique. Neurosurgery. 2013;73(1 Suppl):onsE115-onsE122.

7. Vrionis FD, Cano WG, Heilman CB. Microsurgical anatomy of the infratemporal fossa as viewed laterally and superiorly. Neurosurgery. 1996;39(4):777-786.

8. Sekhar LN, Kalakonda P. Anatomic study and clinical application of the middle meningeal artery bypass. J Neurosurg. 2018;129(5):1213-1221.

9. Vroomen PC, Peeters FL, Bongers V. Cerebral revascularization using the maxillary artery: a review. Neurosurg Rev. 2019;42(1):25-36.

10. Christensen NP, Smith DS, Barnwell SL, Wax MK. Arterial embolization in the management of posterior epistaxis. Otolaryngol Head Neck Surg. 2005;133(5):748-753.

11. Thakar A, Gupta G, Bhalla AS, et al. Adjuvant preoperative embolisation in juvenile nasopharyngeal angiofibroma. Int J Pediatr Otorhinolaryngol. 2012;76(8):1191-1195.

12. Shah A, Giblett S, Kuhn I, et al. Preoperative embolization of intracranial meningiomas: an updated meta-analysis. AJNR Am J Neuroradiol. 2020;41(10):1803-1808.

13. Liu JK, Decker D, Schaefer SD, et al. Zones of attachment and their surgical implications in endoscopic transnasal resection of juvenile nasopharyngeal angiofibromas. Laryngoscope. 2011;121(12):2602-2610.

14. Anxionnat R, Bracard S, Ducrocq X, et al. Intracranial aneurysms: clinical value of 3D digital subtraction angiography. Radiology. 2001;218(3):799-808.

15. Kaufmann TJ, Huston J 3rd, Mandrekar JN, et al. Complications of diagnostic cerebral angiography: evaluation of 19,826 consecutive patients. Radiology. 2007;243(3):812-819.

16. Lell MM, Wildberger JE, Heuschmid M, et al. CT angiography and CT perfusion imaging in clinical practice. Radiologe. 2006;46(6):479-491.

17. Flohr TG, McCollough CH, Bruder H, et al. First performance evaluation of a dual-source CT system. Eur Radiol. 2006;16(2):256-268.

18. Rubin GD. 3-D imaging with MDCT. Eur J Radiol. 2003;45 Suppl 1:S37-S41.

19. Johnson TR, Krauss B, Sedlmair M, et al. Material differentiation by dual energy CT: initial experience. Eur Radiol. 2007;17(6):1510-1517.

20. Anzalone N, Scomazzoni F, Castellano R, et al. Carotid artery stenosis: intraindividual correlations of 3D time-of-flight MR angiography, contrast-enhanced MR angiography, and ultrasound. Radiology. 2005;236(1):204-213.

21. Jin H, et al. Rapid vessel segmentation and reconstruction of head and neck angiograms using 3D convolutional neural network. Nat Commun. 2020;11:4829.

22. Chen T, You W, Zhang L, et al. Automated anatomical labeling of the intracranial arteries via deep learning in computed tomography angiography. Front Physiol. 2024;14:1310357.

23. Wardlaw JM, Chappell FM, Stevenson M, et al. Accurate, practical and cost-effective assessment of carotid stenosis in the UK. Health Technol Assess. 2006;10(30):iii-iv, ix-x, 1-182.

24. Holmstedt CA, Turan TN, Chimowitz MI. Atherosclerotic intracranial arterial stenosis: risk factors, diagnosis, and treatment. Lancet Neurol. 2013;12(11):1106-1114.

25. Taylor-Phillips S, Elze MC, Krupinski EA, et al. Retrospective review of the drop in observer detection performance over time in lesionenriched experimental studies. J Digit Imaging. 2015;28(1):32-40.

26. Doi K. Computer-aided diagnosis in medical imaging: historical review, current status and future potential. Comput Med Imaging Graph. 2007;31(4-5):198-211.

27. Krizhevsky A, Sutskever I, Hinton GE. ImageNet classification with deep convolutional neural networks. Adv Neural Inf Process Syst. 2012;25:1097-1105.

28. Litjens G, Kooi T, Bejnordi BE, et al. A survey on deep learning in medical image analysis. Med Image Anal. 2017;42:60-88.

29. LeCun Y, Bengio Y, Hinton G. Deep learning. Nature. 2015;521(7553):436-444.

30. Zeiler MD, Fergus R. Visualizing and understanding convolutional networks. In: European conference on computer vision. Springer. 2014:818-833.

31. Ronneberger O, Fischer P, Brox T. U-Net: Convolutional Networks for Biomedical Image Segmentation. Medical Image Computing and Computer-Assisted Intervention. Lecture Notes in Computer Science. 2015;9351:234-241.

32. Livne M, et al. A U-Net deep learning framework for high performance vessel segmentation in patients with cerebrovascular disease. Front Neurosci. 2019;13:97.

33. Çiçek Ö, Abdulkadir A, Lienkamp SS, et al. 3D U-Net: Learning Dense Volumetric Segmentation from Sparse Annotation. Medical Image Computing and Computer-Assisted Intervention. 2016;9901:424-432.

34. Zhang Z, Liu Q, Wang Y. Road extraction by deep residual U-Net. IEEE Geosci Remote Sens Lett. 2018;15(5):749-753.

35. Goodfellow I, Pouget-Abadie J, Mirza M, et al. Generative Adversarial Networks. Adv Neural Inf Process Syst. 2014;27:2672-2680.

36. Son J, Park SJ, Jung KH. Towards accurate segmentation of retinal vessels and the optic disc in fundoscopic images with generative adversarial networks. J Digit Imaging. 2019;32(3):499-512.

37. Dosovitskiy A, Beyer L, Kolesnikov A, et al. An image is worth 16x16 words: Transformers for image recognition at scale. arXiv:2010.11929. 2020.

38. Chen J, Lu Y, Yu Q, et al. TransUNet: Transformers make strong encoders for medical image segmentation. arXiv:2102.04306. 2021.

39. Page MJ, McKenzie JE, Bossuyt PM, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;372:n71.

40. Whiting PF, Rutjes AWS, Westwood ME, et al. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med. 2011;155(8):529-536.

41. Mongan J, Moy L, Kahn CE Jr. Checklist for Artificial Intelligence in Medical Imaging (CLAIM): A Guide for Authors and Reviewers. Radiology: Artificial Intelligence. 2020;2(2):e200029.