Problem Statement Intra-operative control of resection borders using systemic cancer labeling and nuclear probes has been proposed as complement for the standard frozen section technique [1]. Advantages are an almost immediate feedback, the gained spatial information (otherwise almost non-existent) and the improved security due to the “closed loop” approach. However, there are issues that have to be solved in order to profit completely of this technology. For example, as radioactive decay and its detection follow a statistical process the readings detected with such probes suffer from a high variance over time. In general their readings are averaged over a certain duration, e.g. from 1 to 3 seconds, in order to provide statistically reliable measurements. This results in a smearing of the readings over an area if the probe is moved and thus in the risk that “invisible” cancer residuals remain undetected. Statistical Correction based on Position To deal with this problem a statistical correction based on tracking of the hand-held detector is introduced. By extending the probe with a tracking system, statistical correction of the readings with a position-?ltering approach is possible. Confidence Calculation In addition the new system allows calculating the confidence level of the estimated probe measurements in the region of interest, which is displayed interactively. This enables guiding the surgeon during the acquisition, making the process less user dependent and more robust. The confidence is calculated from the currently acquired readings and the estimated value using Pearson‘s goodness-of-fit test to calculate a probability. Experiments Our existing tracked beta-probe surface reconstruction system [2] was extended by computation and visualization of the confidence of the reconstruction. The confidence for each pixel of the reconstruction grid (5x5mm² pixel size) is calculated using the simultaneously acquired probe readings and positions. In an experiment we performed 10 scans with and without the visualization of the confidence with 3x3cm² images of a tumor bed phantom with three 5x5mm² residuals (tumor to background 50:10 kBq/ml FDG). Results The results show the superiority of the novel approach. A correlation with the ground truth of 53 - 75% versus 38 - 51% of the standard procedure was achieved. Conclusion Visualizing the confidence during navigated nuclear imaging contributes to a less user-dependent procedure and improved quality of the acquired data. The combination of advanced visualization and guidance with nuclear probes as shown here will thus make it possible to take better advantage of the potential of intra-operative nuclear medicine and open new paths towards a more confident cancer surgery. References [1] Franc, B.L., Mari, C., Johnson, D., Leong, S.P.: The role of a positron- and high-energy gamma photon probe in intraoperative localization of recurrent melanoma. Clin Nucl Med 30(12) (2005) 787–791. [2] Wendler, T., Traub, J., Ziegler, S.I., Navab, N.: Navigated three dimensional beta probe for optimal cancer resection. In: Proc. MICCAI 2006. Volume 1 of LNCS 4190., Copenhagen, Denmark, MICCAI Society, Springer (2006) 561–569.     «

Problem Statement Intra-operative control of resection borders using systemic cancer labeling and nuclear probes has been proposed as complement for the standard frozen section technique [1]. Advantages are an almost immediate feedback, the gained spatial information (otherwise almost non-existent) and the improved security due to the “closed loop” approach. However, there are issues that have to be solved in order to profit completely of this technology. For example, as radioactive decay an...     »