Accelerated 3D-3D rigid registration of echocardiographic images obtained from apical window using particle filter

2025-04-28

AI Summary

Low overlap, high noise, and intensity inhomogeneity in apical-window 3D echocardiographic sequences severely degrade both accuracy and efficiency of 3D–3D rigid registration. To address this, we propose a mask-driven accelerated particle filtering framework. Our method integrates binary left ventricular (LV) mask guidance, dual-path registration (operating jointly on image and mask domains), and iterative refinement of rigid transformations, with parameter estimation performed via accelerated sequential Monte Carlo (SMC). The key innovation is the first introduction of a mask-driven strategy that substantially improves computational efficiency without compromising registration accuracy: achieving an LV Dice coefficient of 0.819 ± 0.045 and accelerating registration by 16.7× over conventional SMC on CPU—enabling clinical real-time performance. Validation on seven 4D cardiac sequences confirms robustness and effectiveness.

Technology Category

• Computer Vision: 3D Computer Vision
• Machine Learning: Multi-instance/Multi-view Learning
• Search and Optimization: Mixed Discrete/Continuous Search

Application Category

• Security and Privacy: Large-scale security measurements
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

The perfect alignment of 3D echocardiographic images captured from various angles has improved image quality and broadened the field of view. This study proposes an accelerated sequential Monte Carlo (SMC) algorithm for 3D-3D rigid registration of transthoracic echocardiographic images with significant and limited overlap taken from apical window that is robust to the noise and intensity variation in ultrasound images. The algorithm estimates the translational and rotational components of the rigid transform through an iterative process and requires an initial approximation of the rotation and translation limits. We perform registration in two ways: the image-based registration computes the transform to align the end-diastolic frame of the apical nonstandard image to the apical standard image and applies the same transform to all frames of the cardiac cycle, whereas the mask-based registration approach uses the binary masks of the left ventricle in the same way. The SMC and exhaustive search (EX) algorithms were evaluated for 4D temporal sequences recorded from 7 volunteers who participated in a study conducted at the Mazankowski Alberta Heart Institute. The evaluations demonstrate that the mask-based approach of the accelerated SMC yielded a Dice score value of 0.819 +/- 0.045 for the left ventricle and gained 16.7x speedup compared to the CPU version of the SMC algorithm.

Problem

• Accelerating 3D-3D rigid registration of echocardiographic images
• Improving alignment of images with limited overlap and noise
• Enhancing speed and accuracy of left ventricle registration

Innovation

• Accelerated SMC algorithm for 3D-3D registration
• Particle filter robust to noise and intensity variation
• Mask-based approach improves speed and accuracy