Department of Computer Science, Johns Hopkins University, Baltimore, 21218, MD, USA.
Department of Neurology, The Johns Hopkins University School of Medicine, Baltimore, 21205, MD, USA.
Int J Comput Assist Radiol Surg. 2022 May;17(5):911-920. doi: 10.1007/s11548-022-02583-6. Epub 2022 Mar 25.
Ultrasound-guided spine interventions often suffer from the insufficient visualization of key anatomical structures due to the complex shapes of the self-shadowing vertebrae. Therefore, we propose an ultrasound imaging paradigm, AutoInFocus (automatic insonification optimization with controlled ultrasound), to improve the key structure visibility.
A phased-array probe is used in conjunction with a motion platform to image a controlled workspace, and the resulting images from multiple insonification angles are combined to reveal the target anatomy. This idea is first evaluated in simulation and then realized as a robotic platform and a miniaturized patch device. A spine phantom (CIRS) and its CT scan were used in the evaluation experiments to quantitatively and qualitatively analyze the advantages of the proposed method over the traditional approach.
We showed in simulation that the proposed system setup increased the visibility of interspinous space boundary, a key feature for lumbar puncture guidance, from 44.13 to 67.73% on average, and the 3D spine surface coverage from 14.31 to 35.87%, compared to traditional imaging setup. We also demonstrated the feasibility of both robotic and patch-based realizations in a spine phantom study.
This work lays the foundation for a new imaging paradigm that leverages redundant and controlled insonification to allow for imaging optimization of the complex vertebrae anatomy, making it possible for high-quality visualization of key anatomies during ultrasound-guided spine interventions.
由于脊柱的自遮挡形状复杂，超声引导下的脊柱介入常常受到关键解剖结构可视化不足的困扰。因此，我们提出了一种超声成像模式，AutoInFocus（带控制超声的自动激发优化），以提高关键结构的可见度。
采用相控阵探头与运动平台相结合来对受控工作空间进行成像，并将来自多个激发角度的图像进行组合以显示目标解剖结构。这一思路首先在模拟中进行了评估，然后实现为一个机器人平台和一个小型贴片设备。在评估实验中，我们使用脊柱模型（CIRS）及其 CT 扫描，对所提出的方法相对于传统方法的优势进行了定量和定性分析。
我们在模拟中表明，与传统成像设置相比，所提出的系统设置将腰椎穿刺引导的关键特征——棘突间空间边界的可见度从平均 44.13%提高到了 67.73%，3D 脊柱表面覆盖率从 14.31%提高到了 35.87%。我们还在脊柱模型研究中展示了机器人和贴片实现的可行性。
这项工作为一种新的成像模式奠定了基础，该模式利用冗余和受控的激发来实现复杂脊柱解剖结构的成像优化，从而有可能在超声引导下的脊柱介入中实现关键解剖结构的高质量可视化。
