GPU implementation of photoacoustic short-lag spatial coherence imaging for improved image-guided interventions.

Eduardo A Gonzalez,Muyinatu A Lediju Bell

doi:10.1117/1.jbo.25.7.077002

Eduardo A Gonzalez, Muyinatu A Lediju Bell

Open Access

https://doi.org/10.1117/1.jbo.25.7.077002

Copy DOI

Abstract

.Significance: Photoacoustic-based visual servoing is a promising technique for surgical tool tip tracking and automated visualization of photoacoustic targets during interventional procedures. However, one outstanding challenge has been the reliability of obtaining segmentations using low-energy light sources that operate within existing laser safety limits.Aim: We developed the first known graphical processing unit (GPU)-based real-time implementation of short-lag spatial coherence (SLSC) beamforming for photoacoustic imaging and applied this real-time algorithm to improve signal segmentation during photoacoustic-based visual servoing with low-energy lasers.Approach: A 1-mm-core-diameter optical fiber was inserted into ex vivo bovine tissue. Photoacoustic-based visual servoing was implemented as the fiber was manually displaced by a translation stage, which provided ground truth measurements of the fiber displacement. GPU-SLSC results were compared with a central processing unit (CPU)-SLSC approach and an amplitude-based delay-and-sum (DAS) beamforming approach. Performance was additionally evaluated with in vivo cardiac data.Results: The GPU-SLSC implementation achieved frame rates up to 41.2 Hz, representing a factor of 348 speedup when compared with offline CPU-SLSC. In addition, GPU-SLSC successfully recovered low-energy signals (i.e., ) with mean ± standard deviation of signal-to-noise ratios of (compared with with conventional DAS beamforming). When energies were lower than the safety limit for skin (i.e., for 900-nm wavelength laser light), the median and interquartile range (IQR) of visual servoing tracking errors obtained with GPU-SLSC were 0.64 and 0.52 mm, respectively (which were lower than the median and IQR obtained with DAS by 1.39 and 8.45 mm, respectively). GPU-SLSC additionally reduced the percentage of failed segmentations when applied to in vivo cardiac data.Conclusions: Results are promising for the use of low-energy, miniaturized lasers to perform GPU-SLSC photoacoustic-based visual servoing in the operating room with laser pulse repetition frequencies as high as 41.2 Hz.

Highlights

Visual servoing[1,2,3] is a promising approach for maintaining visualization of surgical tools during minimally invasive procedures and keeping track of the location of nearby anatomical targets within the body
DC removal was computed by applying one-dimensional convolutions of time-domain kernels, executed independently along the axial dimension. This operation is graphically displayed to the right of the “Remove DC” block in Fig. 2, showing operations at the compute unified device architecture (CUDA) kernel level when transitioning from texture memory to global memory
These processing times are limited by the maximum pulse repetition period (PRP) of any laser, which equals 100 ms for our laser, which has a 10-Hz pulse repetition frequency (PRF)

Summary

Introduction

Visual servoing[1,2,3] is a promising approach for maintaining visualization of surgical tools during minimally invasive procedures and keeping track of the location of nearby anatomical targets within the body. This approach broadly refers to vision-based robot control, and the robot “vision” that we focus on in this paper is provided through photoacoustic images.[4,5] Photoacoustic imaging is achieved by transmitting pulsed light to a structure of interest, which absorbs the light, undergoes thermal expansion, and generates an acoustic response that is received by a conventional ultrasound probe.[6,7,8] This photoacoustic imaging technique was previously demonstrated for multiple applications that require surgery or interventions, such as visualization of brachytherapy seeds,[9,10] intravascular imaging,[11] cardiac catheter visualization,[5] fetal surgeries,[12] prostate surgeries,[13] and endonasal surgeries.[14,15,16] In these applications, structures of interest include blood vessels, nerves, drill tips, and catheter or needle tips.[5,17,18] One or more optical fibers may be coupled to the tool, catheter, or needle tips to transmit the light pulses.[11,19,20,21] Alternatively, a fiber or fiber bundle may be operated independently to provide photoacoustic-based anatomical guidance in the absence of surgical tools.[22,23,24]

Objectives

Methods

Results

Discussion

Conclusion