Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Epicardium Prompt-guided Real-time Cardiac Ultrasound Frame-to-volume Registration

Long Lei, Jun Zhou, Jialun Pei, Baoliang Zhao, Yueming Jin, Yuen-Chun Jeremy Teoh, Jing Qin, Pheng-Ann Heng

TL;DR

This paper tackles real-time registration between a 2D intraoperative ultrasound frame and a 3D preoperative ultrasound volume for guiding cardiac interventions. It introduces CU-Reg, a lightweight end-to-end network that uses epicardium prompt-guided cross-dimensional interaction (PGCA) and a voxel-wise local-global aggregation (VLGA) to fuse 2D sparse and 3D dense features, along with inter-frame discriminative regularization to stabilize adjacent frames. On a CAMUS-derived simulated dataset, CU-Reg achieves higher accuracy and speed than state-of-the-art baselines, with a DistErr around 3.9 mm and runtimes exceeding 35 FPS. The approach enables real-time guidance in cardiac interventions and shows potential for cross-modal ultrasound-CT/MRI fusion applications.

Abstract

A comprehensive guidance view for cardiac interventional surgery can be provided by the real-time fusion of the intraoperative 2D images and preoperative 3D volume based on the ultrasound frame-to-volume registration. However, cardiac ultrasound images are characterized by a low signal-to-noise ratio and small differences between adjacent frames, coupled with significant dimension variations between 2D frames and 3D volumes to be registered, resulting in real-time and accurate cardiac ultrasound frame-to-volume registration being a very challenging task. This paper introduces a lightweight end-to-end Cardiac Ultrasound frame-to-volume Registration network, termed CU-Reg. Specifically, the proposed model leverages epicardium prompt-guided anatomical clues to reinforce the interaction of 2D sparse and 3D dense features, followed by a voxel-wise local-global aggregation of enhanced features, thereby boosting the cross-dimensional matching effectiveness of low-quality ultrasound modalities. We further embed an inter-frame discriminative regularization term within the hybrid supervised learning to increase the distinction between adjacent slices in the same ultrasound volume to ensure registration stability. Experimental results on the reprocessed CAMUS dataset demonstrate that our CU-Reg surpasses existing methods in terms of registration accuracy and efficiency, meeting the guidance requirements of clinical cardiac interventional surgery.

Epicardium Prompt-guided Real-time Cardiac Ultrasound Frame-to-volume Registration

TL;DR

This paper tackles real-time registration between a 2D intraoperative ultrasound frame and a 3D preoperative ultrasound volume for guiding cardiac interventions. It introduces CU-Reg, a lightweight end-to-end network that uses epicardium prompt-guided cross-dimensional interaction (PGCA) and a voxel-wise local-global aggregation (VLGA) to fuse 2D sparse and 3D dense features, along with inter-frame discriminative regularization to stabilize adjacent frames. On a CAMUS-derived simulated dataset, CU-Reg achieves higher accuracy and speed than state-of-the-art baselines, with a DistErr around 3.9 mm and runtimes exceeding 35 FPS. The approach enables real-time guidance in cardiac interventions and shows potential for cross-modal ultrasound-CT/MRI fusion applications.

Abstract

A comprehensive guidance view for cardiac interventional surgery can be provided by the real-time fusion of the intraoperative 2D images and preoperative 3D volume based on the ultrasound frame-to-volume registration. However, cardiac ultrasound images are characterized by a low signal-to-noise ratio and small differences between adjacent frames, coupled with significant dimension variations between 2D frames and 3D volumes to be registered, resulting in real-time and accurate cardiac ultrasound frame-to-volume registration being a very challenging task. This paper introduces a lightweight end-to-end Cardiac Ultrasound frame-to-volume Registration network, termed CU-Reg. Specifically, the proposed model leverages epicardium prompt-guided anatomical clues to reinforce the interaction of 2D sparse and 3D dense features, followed by a voxel-wise local-global aggregation of enhanced features, thereby boosting the cross-dimensional matching effectiveness of low-quality ultrasound modalities. We further embed an inter-frame discriminative regularization term within the hybrid supervised learning to increase the distinction between adjacent slices in the same ultrasound volume to ensure registration stability. Experimental results on the reprocessed CAMUS dataset demonstrate that our CU-Reg surpasses existing methods in terms of registration accuracy and efficiency, meeting the guidance requirements of clinical cardiac interventional surgery.
Paper Structure (14 sections, 7 equations, 3 figures, 1 table)

This paper contains 14 sections, 7 equations, 3 figures, 1 table.

Table of Contents

  1. Introduction
  2. Method
  3. Overview of the proposed CU-Reg
  4. Epicardium Prompt-guided Cross-dimensional Interaction
  5. Voxel-wise Dense Local-Global Aggregation
  6. Hybrid Supervised Learning
  7. Implementation Details
  8. Experiments
  9. Dataset and Evaluation Metrics
  10. Comparison with State-of-the-Art methods
  11. Ablation Study
  12. Conclusion
  13. Acknowledgments.
  14. Disclosure of Interests.

Figures (3)

  • Figure 1: Schematic of cardiac ultrasound frame-to-volume registration.
  • Figure 2: (a) Overview of the proposed CU-Reg, where VLGA is the voxel-wise local-global aggregation; (b) The proposed prompt-guided gated cross-dimensional attention.
  • Figure 3: Qualitative comparison on the registration results of different methods, including predicted slices and their difference heatmaps with the ground truth.