3D/2D Registration of Angiograms using Silhouette-based Differentiable Rendering

TL;DR

This paper introduces a novel differentiable rendering-based method for accurate 3D/2D registration of angiograms, enhancing brain hemodynamics analysis with promising preliminary results on real and synthetic data.

Contribution

It presents a new pose estimation approach using silhouette-based differentiable rendering for 3D/2D angiogram registration, addressing a key challenge in medical imaging.

Findings

Effective registration demonstrated on real and synthetic datasets

Qualitative and quantitative evaluations show promising results

Potential for clinical application in brain hemodynamics analysis

Abstract

We present a method for 3D/2D registration of Digital Subtraction Angiography (DSA) images to provide valuable insight into brain hemodynamics and angioarchitecture. Our approach formulates the registration as a pose estimation problem, leveraging both anteroposterior and lateral DSA views and employing differentiable rendering. Preliminary experiments on real and synthetic datasets demonstrate the effectiveness of our method, with both qualitative and quantitative evaluations highlighting its potential for clinical applications. The code is available at https://github.com/taewoonglee17/TwoViewsDSAReg.