Non-rigid registration based on hierarchical deformation of coronary arteries in CCTA images

Heeryeol Jeong1, Taeyong Park2, Seungwoo Khang1

  • 1School of Computer Science and Engineering , Soongsil University , 369 Sangdo-Ro, Dongjak-Gu, Seoul, 06978 Korea.

Insights

This study introduces a fast and precise method for non-rigid registration of blood vessels in cardiac CT angiography (CTA) images. The technique aids in diagnosing and treating coronary artery diseases with high accuracy and speed.

Area of Science:

  • Medical Imaging
  • Biomedical Engineering
  • Cardiovascular Imaging

Background:

  • Accurate registration of temporal 3D cardiac CT angiography (CTA) images is crucial for diagnosing and treating coronary artery diseases.
  • Non-rigid registration of blood vessels presents challenges due to anatomical variability and dynamic changes.

Purpose of the Study:

  • To develop an accurate and rapid non-rigid registration method for blood vessels in temporal 3D cardiac CTA images.
  • To provide auxiliary information for improved diagnosis and treatment planning of coronary artery diseases.

Main Methods:

  • Global rigid registration of 3D vessel centerlines from temporal CTA images.
  • Point matching and outlier removal using vessel characteristics (thickness, gradient).
  • Hierarchical local non-rigid registration via an energy function.

Main Results:

  • Achieved an average registration error of 0.987 mm.
  • Demonstrated an average execution time of 2.137 seconds.
  • Validated the accuracy and speed of the proposed registration method.

Conclusions:

  • The proposed method offers a rapid and accurate solution for non-rigid registration of cardiac blood vessels.
  • Leveraging vessel characteristics in temporal CTA images enhances registration performance.
  • This technique has significant potential for clinical applications in cardiovascular disease management.