Photoacoustic Image Based Intro-Operative Surgical Guidance System in a da Vinci Surgical Robot Platform
In nerve sparing radical prostatectomy, the surgical guidance system used to track surgical tools helps
surgeons to avoid the intra-operation trauma of nerves and alleviate the post-operative complication.
Instead of using typical electromagnetic or optical trackers [1], we attempt to utilize the photoacoustic
virtual marker (PM) to get the real-time position and orientation about the tool. In previous work, Song
et al. [15] demonstrated the feasibility of this system in ex vivo tissue. Building on Song’s work, we are
further optimizing the whole system. To increase the degree of automation of the system, we integrated
the overall system and implemented a search algorithm for the automatic alignment of PM and TRUS
imaging plane. Figure 2 shows that our original plan used Horn’s method [5] to register the PM coordi-
nates in endoscope camera (EC) image and photoacoustic (PA) image. The main challenge faced by this
paradigm is the error caused by unstable 3D positioning. Also, the alignment is still time-consuming,
even introducing automatic alignment based on a search algorithm. To deal with these drawbacks, we
are developing a novel arc-line registration algorithm that does not require the coordinates of PM in the
EC image. This registration algorithm requires no alignment, eliminates errors caused by inaccurate 3D
coordinates and expands the operating range of the system.
