ntvНаучно-технический вестник информационных технологий, механики и оптикиScientific and Technical Journal of Information Technologies, Mechanics and Optics2226-14942500-0373Университет ИТМО10.17586/2226-1494-2025-25-6-1142-1149ntv-547Research ArticleКОМПЬЮТЕРНЫЕ СИСТЕМЫ И ИНФОРМАЦИОННЫЕ ТЕХНОЛОГИИCOMPUTER SCIENCEАвтоматизированная сегментация коронарных артерий на основе трансформера с доменно-специфической предварительной подготовкойTransformer-based automated coronary artery segmentation with domain-specific pretraininghttps://orcid.org/0009-0003-3545-9719ИсмаилН.IsmailN.

Исмаил Нуар, аспирант

197101; Санкт-Петербург

Nouar Ismail, PhD Student

197101; Saint Petersburg

noauresmail@gmail.comhttps://orcid.org/0000-0002-5483-716XВатьянА. С.VatianA. S.

Александра Сергеевна Ватьян, кандидат технических наук, декан

197101; Санкт-Петербург

sc 57191870868

Alexandra S. Vatian, PhD, Dean

197101; Saint Petersburg

sc 57191870868

alexvatyan@gmail.comhttps://orcid.org/0000-0002-4646-6856БересневА. Д.BeresnevA. D.

Артем Дмитриевич Береснев, кандидат технических наук, заместитель декана

197101; Санкт-Петербург

sc 57202210221

Artem D. Beresnev, PhD, Vice Dean

197101; Saint Petersburg

sc 57202210221

artem.beresnev@itmo.ruhttps://orcid.org/0000-0001-6953-5239ЗубаненкоА. А.ZubanenkoA. A.

Алексей Александрович Зубаненко, генеральный директор ООО, аспирант

197101; 191119; Санкт-Петербург

sc 57215436184

Alexey A. Zubanenko, CEO, Imaging Medical Vision LLC, PhD Student

197101; 191119; Saint Petersburg

sc 57215436184

zubdocmri@gmail.comhttps://orcid.org/0000-0002-1361-6037ГусароваН. Ф.GusarovaN. F.

Наталия Федоровна Гусарова, кандидат технических наук, старший научный сотрудник, доцент

197101; Санкт-Петербург

sc 57162764200

Natalia F. Gusarova, PhD, Senior Researcher, Associate Professor

197101; Saint Petersburg

sc 57162764200

natfed@list.ruhttps://orcid.org/0000-0002-1569-2180МеньковИ. А.Men’kovI. A.

Игорь Анатольевич Меньков, кандидат медицинских наук, заведующий отделением

194044; Санкт-Петербург

Igor A. Men’kov, PhD (Medicine), Head of Department

194044; Saint Petersburg

i.menkov@gmail.comУниверситет ИТМОРоссияITMO UniversityRussian FederationУниверситет ИТМО; ООО ИМВИЖНРоссияITMO University; Imaging Medical Vision LLCRussian FederationВоенно-медицинская академия имени С.М. КироваРоссияThe S.M. Kirov Military Medical AcademyRussian Federation20252312202525611421149Copyright © Исмаил Н., Ватьян А.С., Береснев А.Д., Зубаненко А.А., Гусарова Н.Ф., Меньков И.А., 20252025Исмаил Н., Ватьян А.С., Береснев А.Д., Зубаненко А.А., Гусарова Н.Ф., Меньков И.А.Ismail N., Vatian A.S., Beresnev A.D., Zubanenko A.A., Gusarova N.F., Men’kov I.A.Данная работа распространяется под лицензией Creative Commons Attribution 4.0.This work is licensed under a Creative Commons Attribution 4.0 License.https://ntv.elpub.ru/jour/article/view/547   Введение

   Введение. Автоматизированная сегментация коронарных артерий при компьютерной томографической коронарографии играет важную роль в диагностике и лечении ишемической болезни сердца. Ручная сегментация коронарных артерий требует значительных трудозатрат и сопровождается субъективными погрешностями, что обусловливает необходимость разработки точных и надежных автоматизированных методов сегментации.

   Метод

   Метод. В работе представлен подход на основе глубокой нейронной сети с архитектурой Swin-UNETR, сочетающей преимущества визуальных трансформеров и структуры U-Net. Для повышения точности применена доменно-специфическая стратегия трансферного обучения: модель предварительно обучена на наборе данных ImageCAS, после чего дообучена на специализированном — Automated Segmentation of Normal and Diseased Coronary Arteries (ASOCA) с экспертной разметкой коронарных артерий.

   Основные результаты

   Основные результаты. Оценка точности модели выполнена на 10 тестовых случаях из набора ASOCA. Средний коэффициент Дайса составил 0,8778, а среднее значение 95 % — расстояния Хаусдорфа (HD95) — 11,66 мм. Полученные результаты демонстрируют, что точность предложенного метода находится на уровне ведущих моделей, представленных в официальном рейтинге ASOCA Challenge, и превосходит средний показатель межэкспертной разметки.

   Обсуждение

   Обсуждение. Предложенный метод обеспечивает высокую точность сегментации коронарных артерий. Вместе с тем отмечена вариабельность результатов по HD95, обусловленная отсутствием специализированной постобработки, чувствительностью к мелким сосудам и различиями в характеристиках тестовых данных. В перспективе внедрение методов постобработки, таких как фильтрация связанных компонентов или отслеживание сосудов, а также механизмов пространственного внимания, может повысить точность локализации артериального контура и адаптируемость модели к различным типам данных компьютерной томографии.

   Automated segmentation of coronary arteries in coronary computed tomography angiography plays an important role in the diagnosis and treatment of coronary artery disease. Manual segmentation of coronary arteries requires significant labor costs and is accompanied by subjective errors, which necessitates the development of accurate and reliable automated methods for coronary artery segmentation. The paper presents an approach based on a deep neural network with the Swin-UNETR architecture which combines the advantages of visual transformers and the U-Net structure. To improve the accuracy, a domain-specific transfer learning strategy was used: the model was pre-trained on the ImageCAS dataset, and then further trained on a specialized dataset created for Automated Segmentation of Coronary Arteries (ASOCA) Challenge with expert labeling of coronary arteries. The accuracy of the model was assessed on 10 test Computed Tomography Coronary Angiography cases from the ASOCA dataset. The average Dice coefficient was 0.8778, and the average 95th percentile Hausdorff distance (HD95) was 11.66 mm. The obtained results demonstrate that the accuracy of the proposed method is at the level of the leading models presented in the official ASOCA Challenge rating and exceeds the average inter-rater labeling. The proposed method provides high accuracy of coronary artery segmentation. In the future, the introduction of post-processing methods such as connected component filtering or vessel tracking, and spatial attention mechanisms can improve the accuracy of arterial contour localization and the adaptability of the model to various types of computed tomography data.

коронарные артериикомпьютерная томографическая коронарография с контрастным усилениемглубокое обучениеSwin-UNETRвизуальный трансформерсегментация медицинских изображенийтрансферное обучениеImageCASASOCAcoronary arterycomputed tomography coronary angiography (CTCA)deep learningSwin-UNETRvision transformersmedical image segmentationtransfer learningImageCASASOCAИсследование поддержано грантом Российского научного фонда, № 23-11-00346This work was supported by the Russian Science Foundation, Grant No. 23-11-00346ReferencesRoth G.A., Abate D., Abate K.H., Abay S.M., Abbafati C., Abbasi N., et al. Global, regional, and national age-sex-specific mortality for 282 causes of death in 195 countries and territories, 1980–2017: a systematic analysis for the Global Burden of Disease Study 2017 // The Lancet. 2018. V. 392. N 10159. P. 1736–1788. doi: 10.1016/S0140-6736(18)32203-7Roth G.A., Abate D., Abate K.H., Abay S.M., Abbafati C., Abbasi N., et al. Global, regional, and national age-sex-specific mortality for 282 causes of death in 195 countries and territories, 1980–2017: a systematic analysis for the Global Burden of Disease Study 2017. The Lancet, 2018, vol. 392, no. 10159, pp. 1736–1788. doi: 10.1016/S0140-6736(18)32203-7Montalescot G., Sechtem U., Achenbach S., Andreotti F., Arden C., Budaj A., et al. 2013 ESC guidelines on the management of stable coronary artery disease // European Heart Journal. 2013. V. 34. N 38. P. 2949–3003. doi: 10.1093/eurheartj/eht296Montalescot G., Sechtem U., Achenbach S., Andreotti F., Arden C., Budaj A., et al. 2013 ESC guidelines on the management of stable coronary artery disease. European Heart Journal, 2013, vol. 34, no. 38, pp. 2949–3003. doi: 10.1093/eurheartj/eht296Cury R.C., Leipsic J., Abbara S., Achenbach S., Berman D., Bittencourt M., et al. CAD-RADS™ 2.0 - 2022 Coronary Artery Disease-Reporting and Data System // Journal of Cardiovascular Computed Tomography. 2022. V. 16. N 6. P. 536–557. doi: 10.1016/j.jcct.2022.07.002Cury R.C., Leipsic J., Abbara S., Achenbach S., Berman D., Bittencourt M., et al. CAD-RADS™ 2.0 - 2022 Coronary Artery Disease-Reporting and Data System. Journal of Cardiovascular Computed Tomography, 2022, vol. 16, no. 6, pp. 536–557. doi: 10.1016/j.jcct.2022.07.002Yang S., Chung J., Lesina K., Doh J., Jegere S., Erglis A., et al. Long-term prognostic implications of CT angiography-derived fractional flow reserve: Results from the DISCOVER-FLOW study // Journal of Cardiovascular Computed Tomography. 2024. V. 18. N 3. P. 251–258. doi: 10.1016/j.jcct.2024.01.016Yang S., Chung J., Lesina K., Doh J., Jegere S., Erglis A., et al. Long-term prognostic implications of CT angiography-derived fractional flow reserve: Results from the DISCOVER-FLOW study. Journal of Cardiovascular Computed Tomography, 2024, vol. 18, no. 3, pp. 251–258. doi: 10.1016/j.jcct.2024.01.016Hurreh M., Knickelbine T., Earls J., Aquino M., Idris A., Cheng V. Enhanced cardiac CTA segmental atherosclerosis detection with artificial intelligence quantitative computed tomography (AIQCT) analysis vs conventional reader interpretation // Journal of Cardiovascular Computed Tomography. 2023. V. 17. N 4 Suppl. P. S37-S38. doi: 10.1016/j.jcct.2023.05.093Hurreh M., Knickelbine T., Earls J., Aquino M., Idris A., Cheng V. Enhanced cardiac CTA segmental atherosclerosis detection with artificial intelligence quantitative computed tomography (AIQCT) analysis vs conventional reader interpretation. Journal of Cardiovascular Computed Tomography, 2023, vol. 17, no. 4 suppl., pp. S37-S38. doi: 10.1016/j.jcct.2023.05.093Fedorov A., Beichel R., Kalpathy-Cramer J., Finet J., Fillion-Robin J-C., Pujol S., et al. 3D Slicer as an image computing platform for the Quantitative Imaging Network // Magnetic Resonance Imaging. 2012. V. 30. N 9. P. 1323–1341. doi: 10.1016/j.mri.2012.05.001Fedorov A., Beichel R., Kalpathy-Cramer J., Finet J., Fillion-Robin J-C., Pujol S., et al. 3D Slicer as an image computing platform for the Quantitative Imaging Network. Magnetic Resonance Imaging, 2012, vol. 30, no. 9, pp. 1323–1341. doi: 10.1016/j.mri.2012.05.001Yushkevich P.A., Piven J., Hazlett H.C., Smith R.G., Ho S., Gee J.C., Gerig G. User-guided 3D active contour segmentation of anatomical structures: Significantly improved efficiency and reliability // NeuroImage. 2006. V. 31. N 3. P. 1116–1128. doi: 10.1016/j.neuroimage.2006.01.015Yushkevich P.A., Piven J., Hazlett H.C., Smith R.G., Ho S., Gee J.C., Gerig G. User-guided 3D active contour segmentation of anatomical structures: Significantly improved efficiency and reliability. NeuroImage, 2006, vol. 31, no. 3, pp. 1116–1128. doi: 10.1016/j.neuroimage.2006.01.015Dosovitskiy A., Beyer L., Kolesnikov A., Weissenborn D., Zhai X., Unterthiner T., et al. An image is worth 16×16 words: Transformers for image recognition at scale // arXiv. 2020. arXiv:2010.11929. doi: 10.48550/arXiv.2010.11929Dosovitskiy A., Beyer L., Kolesnikov A., Weissenborn D., Zhai X., Unterthiner T., et al. An image is worth 16×16 words: Transformers for image recognition at scale. arXiv, 2020, arXiv:2010.11929. doi: 10.48550/arXiv.2010.11929Fossan F.E., Müller L.O., Sturdy J., Bråten A.T., Jorgensen A., Wiseth R., Hellevik L.R. Machine learning augmented reduced-order models for FFR-prediction // Computer Methods in Applied Mechanics and Engineering. 2021. V. 384. P. 113892. doi: 10.1016/j.cma.2021.113892Fossan F.E., Müller L.O., Sturdy J., Bråten A.T., Jorgensen A., Wiseth R., Hellevik L.R. Machine learning augmented reduced-order models for FFR-prediction. Computer Methods in Applied Mechanics and Engineering, 2021, vol. 384, pp. 113892. doi: 10.1016/j.cma.2021.113892Gharleghi, R., Adikari D., Ellenberger K., Ooi S.Y., Ellis C., Chen C.M., et al. Automated segmentation of normal and diseased coronary arteries — The ASOCA challenge // Computerized Medical Imaging and Graphics. 2022. V. 97. P. 102049. doi: 10.1016/j.compmedimag.2022.102049Gharleghi, R., Adikari D., Ellenberger K., Ooi S.Y., Ellis C., Chen C.M., et al. Automated segmentation of normal and diseased coronary arteries — The ASOCA challenge. Computerized Medical Imaging and Graphics, 2022, vol. 97, pp. 102049. doi: 10.1016/j.compmedimag.2022.102049Zeng A., Wu C., Lin G., Xie W., Hong J., Huang M.P., et al. ImageCAS: A large-scale dataset and benchmark for coronary artery segmentation based on computed tomography angiography images // Computerized Medical Imaging and Graphics. 2023. V. 109. P. 102287. doi: 10.1016/j.compmedimag.2023.102287Zeng A., Wu C., Lin G., Xie W., Hong J., Huang M.P., et al. ImageCAS: A large-scale dataset and benchmark for coronary artery segmentation based on computed tomography angiography images. Computerized Medical Imaging and Graphics, 2023, vol. 109, pp. 102287. doi: 10.1016/j.compmedimag.2023.102287Çiçek Ö., Abdulkadir A., Lienkamp S.S., Brox T., Ronneberger O. 3D U-Net: Learning dense volumetric segmentation from sparse annotation // Lecture Notes in Computer Science. 2016. V. 9901. P. 424–432. doi: 10.1007/978-3-319-46723-8_49Çiçek Ö., Abdulkadir A., Lienkamp S.S., Brox T., Ronneberger O. 3D U-Net: Learning dense volumetric segmentation from sparse annotation. Lecture Notes in Computer Science, 2016, vol. 9901, pp. 424–432. doi: 10.1007/978-3-319-46723-8_49Chen Y.-C., Lin Y.-C., Wang C.-P., Lee C.-Y., Lee W.-J., Wang T.-D., Chen C.-M. Coronary artery segmentation in cardiac CT Angiography using 3D Multi-channel U-net // arXiv. 2019. arXiv:1907.12246. doi: 10.48550/arXiv.1907.12246Chen Y.-C., Lin Y.-C., Wang C.-P., Lee C.-Y., Lee W.-J., Wang T.-D., Chen C.-M. Coronary artery segmentation in cardiac CT Angiography using 3D Multi-channel U-net. arXiv, 2019, arXiv:1907.12246. doi: 10.48550/arXiv.1907.12246Liang R., Huang X., Ma J., Ma G., Wang K. 3D U-Net with attention and focal loss for coronary tree segmentation // Research Square. Preprint CC BY 4.0 License. 2021. P. 1–17. doi: 10.21203/rs.3.rs-717054/v1Liang R., Huang X., Ma J., Ma G., Wang K. 3D U-Net with attention and focal loss for coronary tree segmentation. Research Square. Preprint CC BY 4.0 License, 2021, pp. 1–17. doi: 10.21203/rs.3.rs-717054/v1Wang Q., Xu L., Wang L., Yang X., Sun Y., Yang B., Greenwald S.E. Automatic coronary artery segmentation of CCTA images using UNet with a local contextual transformer // Frontiers in Physiology. 2023. V. 14. P. 1138257. doi: 10.3389/fphys.2023.1138257Wang Q., Xu L., Wang L., Yang X., Sun Y., Yang B., Greenwald S.E. Automatic coronary artery segmentation of CCTA images using UNet with a local contextual transformer. Frontiers in Physiology, 2023, vol. 14, pp. 1138257. doi: 10.3389/fphys.2023.1138257Zhang L., Yin X., Liu X., Liu Z. Medical image segmentation by combining feature enhancement Swin Transformer and UperNet // Scientific Reports. 2025. V. 15. N 1. P. 14565. doi: 10.1038/s41598-025-97779-6Zhang L., Yin X., Liu X., Liu Z. Medical image segmentation by combining feature enhancement Swin Transformer and UperNet. Scientific Reports, 2025, vol. 15, no. 1, pp. 14565. doi: 10.1038/s41598-025-97779-6Salehi A.W., Khan S., Gupta G., Alabduallah B.I., Almjally A., Alsolai H., et al. A study of CNN and transfer learning in medical imaging: Advantages, challenges, future scope // Sustainability. 2023. V. 15. N 7. P. 5930. doi: 10.3390/su15075930Salehi A.W., Khan S., Gupta G., Alabduallah B.I., Almjally A., Alsolai H., et al. A study of CNN and transfer learning in medical imaging: Advantages, challenges, future scope. Sustainability, 2023, vol. 15, no. 7, pp. 5930. doi: 10.3390/su15075930Matsoukas C., Haslum J.F., Sorkhei M., Soderberg M., Smith K. What makes transfer learning work for medical images: Feature reuse & other factors // Proc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). 2022. P. 9215–9224. doi: 10.1109/CVPR52688.2022.00901Matsoukas C., Haslum J.F., Sorkhei M., Soderberg M., Smith K. What makes transfer learning work for medical images: Feature reuse & other factors. Proc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022, pp. 9215–9224. doi: 10.1109/CVPR52688.2022.00901Yosinski J., Clune O., Bengio Y., Lipson H. How transferable are features in deep neural networks? // Proc. of the 28<sup>th</sup> International Conference on Neural Information Processing Systems. 2014. V. 2. P. 3320–3328.Yosinski J., Clune O., Bengio Y., Lipson H. How transferable are features in deep neural networks? Proc. of the 28<sup>th</sup> International Conference on Neural Information Processing Systems, 2014, vol. 2, pp. 3320–3328.Wang X., Zhang X., Wang G., He J., Li Z., Zhu W., et al. OpenMEDLab: An open-source platform for multi-modality foundation models in medicine // arXiv. 2024. arXiv:2402.18028v2. doi: 10.48550/arXiv.2402.18028Wang X., Zhang X., Wang G., He J., Li Z., Zhu W., et al. OpenMEDLab: An open-source platform for multi-modality foundation models in medicine. arXiv, 2024, arXiv:2402.18028v2. doi: 10.48550/arXiv.2402.18028Tang Y., Yang D., Li W., Roth H.R., Landman B., Xu D., Nath V., Hatamizadeh A. Self-supervised pre-training of Swin Transformers for 3D medical image analysis // Proc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). 2022. P. 20698–20708. doi: 10.1109/cvpr52688.2022.02007Tang Y., Yang D., Li W., Roth H.R., Landman B., Xu D., Nath V., Hatamizadeh A. Self-supervised pre-training of Swin Transformers for 3D medical image analysis. Proc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022, pp. 20698–20708. doi: 10.1109/cvpr52688.2022.02007Yong K.R., Lee S., Uppaluri A., Tan M.O., Ng J.J., Venkatesh S.H., et al. How to report a coronary CT angiogram: Step-by-Step guide for the novice reader // RadioGraphics. 2025. V. 45. N 7. P. e250015. doi: 10.1148/rg.250015Yong K.R., Lee S., Uppaluri A., Tan M.O., Ng J.J., Venkatesh S.H., et al. How to report a coronary CT angiogram: Step-by-Step guide for the novice reader. RadioGraphics, 2025, vol. 45, no. 7, pp. e250015. doi: 10.1148/rg.250015Gharleghi R., Adikari D., Ellenberger K., Webster M., Ellis C., Sowmya A., Ooi S., Beier S. Annotated computed tomography coronary angiogram images and associated data of normal and diseased arteries // Scientific Data. 2023. V. 10. N 1. P. 128. doi: 10.1038/s41597-023-02016-2Gharleghi R., Adikari D., Ellenberger K., Webster M., Ellis C., Sowmya A., Ooi S., Beier S. Annotated computed tomography coronary angiogram images and associated data of normal and diseased arteries. Scientific Data, 2023, vol. 10, no. 1, pp. 128. doi: 10.1038/s41597-023-02016-2Kirişli H.A., Schaap M., Metz C.T., Dharampal A.S., Meijboom W.B., Papadopoulou S.L., et al. Standardized evaluation framework for evaluating coronary artery stenosis detection, stenosis quantification and lumen segmentation algorithms in computed tomography angiography // Medical Image Analysis. 2013. V. 17. N 8. P. 859–876. doi: 10.1016/j.media.2013.05.007Kirişli H.A., Schaap M., Metz C.T., Dharampal A.S., Meijboom W.B., Papadopoulou S.L., et al. Standardized evaluation framework for evaluating coronary artery stenosis detection, stenosis quantification and lumen segmentation algorithms in computed tomography angiography. Medical Image Analysis, 2013, vol. 17, no. 8, pp. 859–876. doi: 10.1016/j.media.2013.05.007

The authors declare that there are no conflicts of interest present.