Wave-Cam: A novel micro-radar imaging array for non-rigid motion estimation in hybrid medical imaging

摘要 在医学成像期间，患者运动是不可避免的，并且可能导致显著的图像退化。心脏 疾病和癌症几乎占美国所有死亡人数的一半， 对于这些疾病的诊断、治疗监测和手术规划， 运动会产生重大影响。此外，在多模态成像中， 连续采集之间的患者运动使模态间对准失真，降低了成像的效用。 现有的运动校正策略被分类为：1）通过利用成像数据的数据驱动或2）使用 成像系统外部的设备，其通过以下方式估计内部感兴趣体积（VOI）的运动： 跟踪替代物的运动，例如患者的身体表面。外部运动跟踪（EMT）是一种 数据驱动方法的流行替代方案，因为它不受模态特性（如噪声水平）的影响 和放射性示踪剂分布，并可用于多模式系统中的跟踪。大多数急救摄像头 使用被患者反射回的可见光或近红外电磁辐射主动照射患者 服装.服装对这种类型的辐射的不透明性阻止了对患者实际身体的跟踪 表面，从而在估计运动时增加了一定程度的不确定性。如果病人有病， 当前的EMT仅跟踪自身充当患者身体的替代物的服装或标记 表面，这等同于跟踪VOI的替代到另一替代。我们的前提是病人 服装对EMT构成了难以逾越的障碍，限制了他们的运动跟踪和补偿 在临床成像中的表现。因此，我们建议开发一种新的EMT，它利用了 电磁频谱，衣服是透明的，从而克服了这一限制，反过来， 改变在临床环境中执行运动跟踪的方式。

Abstract During medical imaging, patient motion is unavoidable and can result in significant image degradation. Heart disease and cancer account for almost half of all deaths in the U.S. and considering the prevalence of imaging for diagnosis, therapeutic monitoring, and surgical planning for these diseases, degraded images due to patient motion can have significant implications. Furthermore, in multi-modal imaging the potential for misalignment from patient motion between sequential acquisitions distorts intermodality alignment decreasing the utility of imaging. Existing strategies for motion correction are classified as either: 1) data-driven by utilizing imaging data or 2) use of a device external to the imaging system(s) that estimates motion of internal volumes of interest (VOI) by tracking the motion of a surrogate, such as the patient’s body surface. External motion tracking (EMT) is a popular alternative to data-driven approaches as it is unaffected by modality characteristics, such noise levels and radiotracer distribution and can be used for tracking in multi-modal systems. The majority of EMT cameras actively illuminate patients with visible or near infrared electromagnetic radiation that is reflected back by patient garments. The opaqueness of garments to this type of radiation prevents tracking of the patient’s actual body surface, thereby adding a level of uncertainty when estimating the motion. Thus, if the patient is clothed, then current EMTs only track garments or markers which are themselves acting as surrogates to the patient’s body surface, which is tantamount to tracking a surrogate to another surrogate to the VOI. Our premise is that patient garments pose an unsurmountable barrier to EMTs that limit their motion tracking and compensation performance in clinical imaging. As such, we propose to develop a novel EMT that utilizes a part of the electromagnetic spectrum for which clothing is transparent, thus overcoming this limitation, and in turn transforming the way motion tracking is performed in clinical settings.