Computer Aided Medical Procedures, Technical University of Munich, Munich, Germany.
Trinity College Dublin, Dublin, Ireland.
Int J Comput Assist Radiol Surg. 2018 May;13(5):703-711. doi: 10.1007/s11548-018-1723-9. Epub 2018 Mar 15.
Optoacoustic imaging provides high spatial resolution and the possibility to image specific functional parameters in real-time, therefore positioning itself as a promising modality for various applications. However, despite these advantages, the applicability of real-time optoacoustic imaging is generally limited due to a relatively small field of view.
With this work, we aim at presenting a path towards panoramic optoacoustic tomographic imaging without requiring additional sensors or position trackers. We propose a two-step seamless stitching method for the compounding of multiple datasets acquired with a real-time 3D optoacoustic imaging system within a panoramic scan. The employed workflow is specifically tailored to the image properties and respective challenges.
A comparison of the presented alignment on in-vivo data shows a mean error of [Formula: see text] compared to ground truth tracking data. The presented compounding scheme integrates the physical resolution of optoacoustic data and hence can provide improved contrast in comparison with other compounding approaches based on addition or averaging.
The proposed method can produce optoacoustic volumes with an enlarged field of view and improved quality compared to current methods in optoacoustic imaging. However, our study also shows challenges for panoramic scans. In this view, we discuss relevant properties, challenges, and opportunities and present an evaluation of the performance of the presented approach with different input data.
光声成像是一种很有前途的成像方式,具有高空间分辨率和实时成像特定功能参数的可能性。然而,尽管有这些优势,实时光声成像的适用性通常由于视场较小而受到限制。
本研究旨在提出一种无需额外传感器或位置跟踪器即可实现全景光声断层成像的方法。我们提出了一种两步无缝拼接方法,用于在全景扫描中组合使用实时 3D 光声成像系统采集的多个数据集。所采用的工作流程是专门针对图像特性和各自的挑战而定制的。
对体内数据的对齐比较表明,与地面实况跟踪数据相比,平均误差为[公式:见正文]。所提出的合成方案集成了光声数据的物理分辨率,因此与基于相加或平均的其他合成方法相比,可以提供更好的对比度。
与光声成像中的现有方法相比,所提出的方法可以产生具有更大视场和更高质量的光声体积。然而,我们的研究也显示出全景扫描的挑战。在这方面,我们讨论了相关的性质、挑战和机遇,并对不同输入数据的提出的方法的性能进行了评估。