SlingBAG: point cloud-based iterative algorithm for large-scale 3D photoacoustic imaging

Abstract

Large-scale 3D photoacoustic imaging has become increasingly important for both clinical and pre-clinical applications. Limited by cost and system complexity, only systems with sparsely-distributed sensors can be widely implemented, which necessitates advanced reconstruction algorithms to reduce artifacts. However, the high computing memory and time consumption of traditional iterative reconstruction (IR) algorithms is practically unacceptable for large-scale 3D photoacoustic imaging. Here, we propose a point cloud-based IR algorithm that reduces memory consumption by several orders, wherein the 3D photoacoustic scene is modeled as a series of Gaussian-distributed spherical sources stored in form of point cloud. During the IR process, not only are properties of each Gaussian source, including its peak intensity (initial pressure value), standard deviation (size) and mean (position) continuously optimized, but also each Gaussian source itself adaptively undergoes destroying, splitting, and duplication along the gradient direction. This method, named SlingBAG, the sliding Gaussian ball adaptive growth algorithm, enables high-quality large-scale 3D photoacoustic reconstruction with fast iteration and extremely low memory usage. We validated the SlingBAG algorithm in both simulation study and in vivo animal experiments.

Researchers present SlingBAG, an iterative reconstruction algorithm for large-scale 3D photoacoustic imaging. It uses an adaptive point cloud model to achieve high-quality imaging from sparse data, notably cutting cost in both memory and time.