Neural Network Vessel Lumen Regression for Automated Lumen Cross-Section Segmentation in Cardiovascular Image-Based Modeling

Cardiovasc Eng Technol. 2020 Dec;11(6):621-635. doi: 10.1007/s13239-020-00497-5. Epub 2020 Nov 11.


Purpose: We accelerate a pathline-based cardiovascular model building method by training machine learning models to directly predict vessel lumen surface points from computed tomography (CT) and magnetic resonance (MR) medical image data.

Methods: We formulate vessel lumen detection as a regression task using a polar coordiantes representation.

Results: Neural networks trained with our regression formulation allow predictions to be made with significantly higher accuracy than existing methods that identify the vessel lumen through binary pixel classification. The regression formulation enables machine learning models to be trained end-to-end for vessel lumen detection without post-processing steps that reduce accuracy.

Conclusion: By employing our models in a pathline-based cardiovascular model building pipeline we substantially reduce the manual segmentation effort required to build accurate cardiovascular models, and reduce the overall time required to perform patient-specific cardiovascular simulations. While our method is applied here for cardiovascular model building it is generally applicable to segmentation of tree-like and tubular structures from image data.

Keywords: Cardiovascular modeling; Cardiovascular simulation; Convolutional neural networks; Patient-specific modeling; SimVascular.

Publication types

  • Research Support, N.I.H., Extramural
  • Research Support, Non-U.S. Gov't
  • Research Support, U.S. Gov't, Non-P.H.S.

MeSH terms

  • Automation
  • Blood Vessels / diagnostic imaging*
  • Computed Tomography Angiography*
  • Diagnosis, Computer-Assisted*
  • Humans
  • Magnetic Resonance Angiography*
  • Models, Cardiovascular*
  • Neural Networks, Computer*
  • Patient-Specific Modeling*
  • Predictive Value of Tests
  • Radiographic Image Interpretation, Computer-Assisted*
  • Reproducibility of Results