Figure 1: Current Setup of the Photoacoutic Registration System Interventional Photoacoustic Registration 600.446: Computer Integrated Surgery II Spring 2011 Saurabh Vyas, Steven Su & Robert Kim Mentors: Dr. Emad M. Boctor, Dr. Russell H. Taylor & Dr. Jin U. Kang Introduction When matter is exposed to high frequency pulsed laser it causes pressure variations in the incident molecules. These variations emit sound waves which can be recorded using an ultrasound transducer. This is known as the photoacoustic effect. This project involves developing a new and novel technology for surgical navigation using photoacoustic imaging. It promises to replace current EM and Optical tracking modalities. The group developed a powerful software package to compute the rigid registration between Ultrasound (US) and Stereo Camera (SC) coordinate systems. The group designed experiments for 532 nm and 1064 nm wavelength lasers to collect real-time US and SC data. The real-time US and SC data was used to verify the forward & reverse registrations between coordinate systems, hence resulting in a model for the registration between US and SC coordinates. Outcomes and Results The group was able to design the Photoacoustic Registration System. Figure 2 below shows the beam formed image and its centroid computed with image processing algorithms as well as the same point computed from the SC image using the calculated registration. The registration was verified by picking a sample point in the SC system, and then applying the rigid transformation and measuring the error between the predicted transformed point and the actual point in the US system. Figure 3 below summarizes the actual coordinates of the centroid, the predicted coordinates, and the difference between actual and predicted (i.e. mm-accuracy of the registration) This creation offers another (novel) surgical navigation modality which offers several benefits over current systems. Figure 2: Beam formed image showing the actual centroid (green) computed using image processing algorithms as well as the centroid predicted by the rigid transformation (red) from SC to US plotted together. The error in this trial was 0.384 mm The Problem Problems with current surgical navigation systems: Large and intrusive (large base stations, wires, etc.) Difficult to manipulate and modify Steep learning curve Extensive setup procedures Ergonomically difficult for surgeons Prone to error (i.e. line of sight error or interference) Expensive The Solution The group created a new method which involved firing a pulsed as well as a continuous wave laser at the same target of interest. An US image as well as a SC image was obtained while the target was exposed to the laser sources (pulsed and continuous). The US image and well as the SC image contained marker points corresponding to the laser activity (i.e. the SC image could “see” the continuous wave laser spots while the points in the US image were obtained photoacoustically). A registration was computed between the points seen in the US image and the points seen in the SC image. This registration was then verified by transforming a set of sample points in one coordinate system into the another, and then computing the error between the predicted transformed points and the actual points in that system. Figure 3: Table comparing the actual coordinates of sample points with that of coordinates predicted by the rigid transformation. The error between the data as well as the average error is also reported. Future Work Purchase and design a more complex “caged optics laser system” setup to obtain a less scattered laser beam. Design a prototype of the iPASS Platform. Lessons Learned Crash course on medical imaging and applications Experience with photonics, lasers and US technology Extensive experience with I mage processing algorithms Credits All authors contributed equally to the final execution and implementation of the project. Publications N. Kuo, et al. "Photoacoustic imaging of prostate brachytherapy seeds in ex vivo prostate", Proc. SPIE 7964, 796409 [2011] Xu et al. Photoacoustic Imaging in Biomedicine. Review of Scientific Instruments. [2006] C. G. A. Hoelen et al. 3-Dimensional Photoacoustic Imaging of Blood Vessels in Tissue. Optic Letters. [1998] Support by and Acknowledgements Thanks to Dr. Russ Taylor and Dr. Emad Boctor for their support and encouragement with this project. Special thanks to Nathanael Kuo for consistently taking time out of his schedule every week to help us. Figure 1: Current Setup of the Photoacoutic Registration System Engineering Research Center for Computer Integrated Surgical Systems and Technology