Robotic Repositioning of Photoacoustic Tomography for Reproducible Long-Term Monitoring of Peripheral Artery In Vivo

Abstract

This paper presents a robotic arm assisted photoacoustic tomography imager that employs a direct visual servoing-enabled repositioning strategy designed for reproducible surveillance of peripheral arterial disease (PAD) progression. The system ensures precise repositioning of the cross-sectional location of the vessel of interest, allowing consistent measurements to track PAD progression and response to treatment. This strategy overcomes the difficulty of obtaining cross-sectional images of the same vessel at different monitoring intervals due to operator variability. In addition, the acquisition of high-resolution, high-contrast vessel images helps to clearly delineate long-term changes in the vasculature. To evaluate the effectiveness of the proposed experimental configurations and algorithms, we performed two sets of vascular phantom experiments and one set of servo-imaging experiments on human lower extremity vessels. The experimental results show 100% similarity for all three sets of grid segmentation comparisons, while the pixel-wise comparison similarity is 99.3% for the first set, 97.4% for the second set, and 98.9% for the in vivo experimental set. These results are important for monitoring the progression of PAD and predicting the risk of cardiovascular disease.