Registration of Magnetic Resonance and 3D Ultrasound for Renal Intervention

Abstract

In this paper, we propose an accurate, robust and swift registration method of magnetic resonance (MR) and 3D ultrasound for percutaneous renal intervention. We refine the traditional ICP algorithm with a pre-registration Procrustes initial alignment to improve the accuracy and efficiency. We propose the orthogonal feature extraction principle to increase the robustness while reducing the human interaction required during the registration. The accuracy, robustness and efficiency performance of the proposed registration method is examined via experiment on real human data.