Patch Fusion: A Novel Ultrafast Multi-Frequency Ultrasound Fusion Imaging Method for Pedicle Screw Navigation.

Abstract

Intraosseous ultrasound imaging is valuable for guiding pedicle screw placement in surgery. However, single-frequency ultrasound, whether low or high, often fails to provide both adequate imaging resolution and depth simultaneously. To address this limitation, we introduce a novel ultrafast multi-frequency ultrasound patch fusion imaging method for pedicle screw navigation. This approach combines the strengths of both high-frequency and low-frequency ultrasound images, greatly enhancing the detail and clarity of the resulting images while significantly reducing the time required for image fusion. We validated our method through simulation and ex-vivo experiments, using metrics such as Information Entropy (IE), Spatial Frequency (SF), and Average Gradient (AG) to assess the quality of the fused images. We also recorded the algorithm's execution time. The results demonstrate that our fusion method substantially improves image richness and clarity, enabling a more comprehensive and accurate assessment of the pedicle screw track. Importantly, it also reduces fusion time compared to previous methods, making real-time clinical multi-frequency ultrasound fusion imaging a viable possibility. The in-vivo experimental results of the sheep spinal pedicle screw track further demonstrate the capabilities of the patch fusion method in visualizing the internal conditions of the pedicle screw track and meeting the requirements for real-time fusion imaging. The proposed approach offers substantial support in surgical real-time navigation and ongoing monitoring within the domains of orthopedics and surgery.