Korean Circulation Journal

The Korean Society of Cardiology (대한심장학회)
권호 표지
DOI QR코드

DOI QR Code

Deep Learning-Based Lumen and Vessel Segmentation of Intravascular Ultrasound Images in Coronary Artery Disease

  • Gyu-Jun Jeong (Biomedical Engineering Research Center, Asan Institute for Life Sciences) ;
  • Gaeun Lee (Biomedical Engineering Research Center, Asan Institute for Life Sciences) ;
  • June-Goo Lee (Biomedical Engineering Research Center, Asan Institute for Life Sciences) ;
  • Soo-Jin Kang (Department of Cardiology, Asan Medical Center, University of Ulsan College of Medicine)
  • Received : 2023.06.14
  • Accepted : 2023.09.19
  • Published : 2024.01.01
Download PDF
⟨ Previous Next ⟩

Abstract

Background and Objectives: Intravascular ultrasound (IVUS) evaluation of coronary artery morphology is based on the lumen and vessel segmentation. This study aimed to develop an automatic segmentation algorithm and validate the performances for measuring quantitative IVUS parameters. Methods: A total of 1,063 patients were randomly assigned, with a ratio of 4:1 to the training and test sets. The independent data set of 111 IVUS pullbacks was obtained to assess the vessel-level performance. The lumen and external elastic membrane (EEM) boundaries were labeled manually in every IVUS frame with a 0.2-mm interval. The Efficient-UNet was utilized for the automatic segmentation of IVUS images. Results: At the frame-level, Efficient-UNet showed a high dice similarity coefficient (DSC, 0.93±0.05) and Jaccard index (JI, 0.87±0.08) for lumen segmentation, and demonstrated a high DSC (0.97±0.03) and JI (0.94±0.04) for EEM segmentation. At the vessel-level, there were close correlations between model-derived vs. experts-measured IVUS parameters; minimal lumen image area (r=0.92), EEM area (r=0.88), lumen volume (r=0.99) and plaque volume (r=0.95). The agreement between model-derived vs. expert-measured minimal lumen area was similarly excellent compared to the experts' agreement. The model-based lumen and EEM segmentation for a 20-mm lesion segment required 13.2 seconds, whereas manual segmentation with a 0.2-mm interval by an expert took 187.5 minutes on average. Conclusions: The deep learning models can accurately and quickly delineate vascular geometry. The artificial intelligence-based methodology may support clinicians' decision-making by real-time application in the catheterization laboratory.

Keywords

Acknowledgement

This study was supported by grants from the Ministry of Science and ICT (NRF2021R1A2C2006831) and the Asan Institute for Life Sciences, Asan Medical Center, Seoul, Republic of Korea (grant Nos. 2021IP0071-1 and 2019IE7053).

References

  1. Fujii K, Carlier SG, Mintz GS, et al. Stent underexpansion and residual reference segment stenosis are related to stent thrombosis after sirolimus-eluting stent implantation: an intravascular ultrasound study. J Am Coll Cardiol 2005;45:995-8.
  2. Okabe T, Mintz GS, Buch AN, et al. Intravascular ultrasound parameters associated with stent thrombosis after drug-eluting stent deployment. Am J Cardiol 2007;100:615-20.
  3. Liu X, Doi H, Maehara A, et al. A volumetric intravascular ultrasound comparison of early drug-eluting stent thrombosis versus restenosis. JACC Cardiovasc Interv 2009;2:428-34.
  4. Stone GW, Maehara A, Lansky AJ, et al.. A prospective natural-history study of coronary atherosclerosis. N Engl J Med 2011;364:226-35.
  5. Bajaj R, Huang X, Kilic Y, et al. Advanced deep learning methodology for accurate, real-time segmentation of high-resolution intravascular ultrasound images. Int J Cardiol 2021;339:185-91.
  6. Zhu F, Gao Z, Zhao C, et al. A deep learning-based method to extract lumen and media-adventitia in intravascular ultrasound images. Ultrason Imaging 2022;44:191-203.
  7. Szarski M, Chauhan S. Improved real-time segmentation of intravascular ultrasound images using coordinate-aware fully convolutional networks. Comput Med Imaging Graph 2021;91:101955.
  8. Destrempes F, Roy Cardinal MH, Allard L, Tardif JC, Cloutier G. Segmentation method of intravascular ultrasound images of human coronary arteries. Comput Med Imaging Graph 2014;38:91-103.
  9. Brunenberg E, Pujol O, ter Haar Romeny B, Radeva P. Automatic IVUS segmentation of atherosclerotic plaque with stop & go snake. Med Image Comput Comput Assist Interv 2006;9:9-16.
  10. Mendizabal-Ruiz EG, Rivera M, Kakadiaris IA. Segmentation of the luminal border in intravascular ultrasound B-mode images using a probabilistic approach. Med Image Anal 2013;17:649-70.
  11. Yang J, Faraji M, Basu A. Robust segmentation of arterial walls in intravascular ultrasound images using Dual Path U-Net. Ultrasonics 2019;96:24-33.
  12. Nishi T, Yamashita R, Imura S, et al. Deep learning-based intravascular ultrasound segmentation for the assessment of coronary artery disease. Int J Cardiol 2021;333:55-9.