Three-dimensional Elastic Augmented-reality Robot-assisted Radical Prostatectomy Using Hyperaccuracy Three-dimensional Reconstruction Technology: A Step Further in the Identification of Capsular Involvement

Abstract

Background: In prostate cancer (PCa) surgical procedures, in order to maximize potency recovery, a nerve-sparing (NS) procedure is preferred. However, cancer abutting or focally extending beyond the prostate capsule increases the risk of a positive surgical margin.

Objective: To evaluate the accuracy of our new three-dimensional (3D) elastic augmented-reality (AR) system in identifying capsular involvement (CI) location of PCa during the NS phase of robot-assisted radical prostatectomy (RARP). Secondarily, the accuracy of this technology was compared with two-dimensional (2D)-based cognitive procedures.

Design, setting, and participants: A prospective study, enrolling 40 patients with PCa undergoing RARP at our center, from May to October 2018.

Surgical procedure: Patients underwent 3D AR RARP or, in case of unavailability of this technology, 2D cognitive RARP. In all patients, total anatomical reconstruction was used.

Measurements: Clinical data were collected. In order to compare the two groups, nonparametric Mann-Whitney and chi-square tests were performed. A metallic clip was placed at the level of suspicious CI on the basis of images given by the 3D AR or magnetic resonance imaging (MRI) report. The pathological analysis evaluated the presence of tumor at the level of the clip.

Results and limitations: Twenty patients were enrolled in each group. Focusing on the 3D AR group at macroscopic evaluation, the metallic clip was placed at the tumor and capsular bulging in all cases. At microscopic assessment, cancer presence was confirmed in the suspicious area in 95.4% of the cases. Moreover, CI was correctly identified in 100.0% of the cases, thanks to the 3D image overlap. These results were compared with the 2D MRI cognitive group, showing, at microscopic analysis, statistically significant superiority of the 3D AR group in CI detection during the NS phase (100% vs 47.0%; p<0.05). The main limitation of this technique is that the segmentation and overlapping of the images are performed manually.

Conclusions: Our findings suggest that, with the introduction of the elastic 3D virtual models, prostate deformation is correctly simulated during surgery and lesion location is correctly identified, even in dynamic reality with a subsequent potential reduction of positive surgical margin rate and, in the meantime, maximization of functional outcomes.

Patient summary: On the basis of our findings, the three-dimensional elastic augmented-reality technology seems to help the surgeon in lesion location identification even in a dynamic phase of the intervention, optimizing the oncological outcomes.

Keywords: Augmented reality; Hyperaccuracy three-dimensional reconstruction; Image-guided surgery; Prostatectomy; Robotics; Three-dimensional reconstruction.