Insonification Angle-based Ultrasound Volume Reconstruction for Spine Intervention

TL;DR

This paper introduces a wearable multi-angle ultrasound system with a novel reconstruction algorithm that enhances spine bone surface visualization, aiding in more accurate and safer spine interventions.

Contribution

The study presents a new multi-angle ultrasound imaging device combined with an angle-weighted reconstruction algorithm for improved spine visualization during interventions.

Findings

Enhanced bone surface contrast in ultrasound images

Improved visualization clarity for clinicians

Validated on spine phantom data

Abstract

Ultrasound-guided spine interventions, such as lumbar-puncture procedures, often suffer from the reduced visibility of key anatomical features such as the inter-spinous process space, due to the complex shape of the self-shadowing vertebra. Therefore, we propose to design a wearable 3D ultrasound device capable of imaging the vertebra from multiple insonification angles to improve the 3D bone surface visualization for interventional guidance. In this work, we aim to equip the imaging platform with a reconstruction algorithm taking advantage of the redundant ultrasound beam angles. Specifically, we try to weight each beam's contribution for the same reconstructed voxel during the reconstruction process based on its incidence angle to the estimated bone surface. To validate our approach, we acquired multi-angle ultrasound image data on a spine phantom with a tracked phased array transducer. The results show that with the proposed method the bone surface contrast can be significantly enhanced, providing clearer visual guidance for the clinician to perform spine intervention.