Application of a three-dimensional visualization model in intraoperative guidance of percutaneous nephrolithotomy.

ABSTRACT

OBJECTIVES: To establish a three-dimensional visualization model of percutaneous nephrolithotomy, apply it to guiding intraoperative puncture in a mixed reality environment, and evaluate its accuracy and clinical value. METHODS: Patients with percutaneous nephrolithotomy indications were prospectively divided into three-dimensional group and control group with a ratio of 12. For patients in three-dimensional group, positioning markers were pasted on the skin and enhanced computed tomography scanning was performed in the prone position. Holographic three-dimensional models were made and puncture routes were planned before operation. During the operation, the three-dimensional model was displayed through HoloLens glass and visually registered with the patient's body. Puncture of the target renal calyx was performed under three-dimensional-image guiding and ultrasonic monitoring. Patients in the control group underwent routine percutaneous nephrolithotomy in the prone position under the monitoring of B-ultrasound. Deviation distance of the kidney, puncture time, puncture attempts, channel coincidence rate, stone clearance rate, and postoperative complications were assessed. RESULTS: Twenty-one and 40 patients were enrolled in three-dimensional and control group, respectively. For three-dimensional group, the average deviation between virtual and real kidney was 3.1 ± 2.9 mm. All punctures were performed according to preoperative planning. Compared with the control group, the three-dimensional group had shorter puncture time (8.9 ± 3.3 vs 14.5 ± 6.1 min, P < 0.001), fewer puncture attempts (1.4 ± 0.6 vs 2.2 ± 1.5, P = 0.009), and might also have a better performance in stone clearance rate (90.5% vs 72.5%, P = 0.19) and postoperative complications (P = 0.074). CONCLUSIONS: The percutaneous nephrolithotomy three-dimensional model manifested acceptable accuracy and good value for guiding puncture in a mixed reality environment.