Abstract
Purpose :
The table-top setting of OCT angiography (OCTA) impedes its application in pediatrics, acute care, and intraoperative monitoring. We developed an OCTA imaging probe with two handheld operation modes for versatile clinical scenarios. The imaging optics was incorporated with a 3-axis gimbal for minimizing the motion induced by the handheld operation.
Methods :
We built a spectral-domain imaging system for OCTA acquisition. We employed the Cobra-S spectrometer, which integrates the state-of-the-art CMOS line camera with an A-line rate of 250 kHz. To compensate for the signal degeneration caused by a relatively low bit-depth (10 bit) at this speed, we developed a deep-learning-based enhancement method. For optimizing the handheld alignment process, we employed an electrically tunable lens for fast automatic focusing. Gimbal stabilizer has been widely used in live streaming and on-site photography for improving imaging quality. Here we introduced this technology into the handheld OCTA imaging. We built a handheld sample arm with the optomechanical design adapted with a motorized 3-axis gimbal stabilizer. Besides, we proposed to leverage adversarial deep learning to reduce OCTA sampling rate while preserving vessel visibility, which could further shorten acquisition time thus minimizing the influence of motion.
Results :
For the younger children and disabled people who are able to sit or stand but unable to perform the table-top OCTA acquisition, we could use the sitting/standing mode in Fig. 1(a). Inside the 3D-printed casing (white) are the sample-arm optics and galvo scanner, which are mounted on the gimbal stabilizer (black). For neonates, infants, and intraoperative subjects, the lying operation mode in Fig. 1(b) should be adopted. To verify the efficacy of the motion stabilization, we examined how the directionality of the probe beam changes over time (in 1 minute). The results are shown in Fig. 2. The gray points are the irradiation positions over time. With the gimbal stabilizer, the beam stability is significantly improved.
Conclusions :
The proposed motion-stabilized dual-mode handheld OCTA could benefit the population groups which are unable to perform the table-top OCTA imaging.
This is a 2020 ARVO Annual Meeting abstract.