AUA 2018: Endoscopic Stone Measurement During Ureteroscopy
As discussed by Dr. Joice, the size of the object viewed through a ureteroscope depends on the distance from the end of the scope. The scale of the ureteroscopic image and the distance of the endoscopic instrument have a linear relationship, thus researchers used these parameters to develop the software in C++. To test the efficacy and efficiency of the developed software, two separate ureteroscopes were used to take 10 images of an 8x6 checkerboard with known dimensions. The image, which is naturally distorted by the scope, was then undistorted with a general calibration technique from the novel software. The objects were then measured based on image scale. Then, the software was tested during standard of care URS within 5 patients with renal and ureteral stones using a LithoVue ureteroscope, then videos from the surgery were saved. 30 stone fragments were collected from these cases and then subsequently measured. “Intraoperative” stone measurements were collected by analyzing the saved video after the procedure. Software error, accuracy, and precision were then calculated.
Through these procedures, maximum errors of measurement were 0.53 and 0.59 mm with an accuracy of 0.17 mm and precision of 0.15 mm. Stone size and composition were not correlated with error measurements. Through further testing with fifty mock stone fragments sized from 2 to 6 mm, it was determined that the maximum stone dimension of less than 3.85 mm led to a 100% specificity within the experimental software group. This value was much greater than the six independent observers (urology residents, fellows, and attendings) who were asked to attempt to evaluate fragment images and determine whether the stone would be able to fit through a 12/14F ureteral access sheath.
In closing, Dr. Joice described that his software can accurately and reliably provide stone measurement during ureteroscopy. The software has been shown to efficiently select stones that will fit through an access sheath with high specificity, while also outperforming trained urologists.
Presented by: Gregory Joice, MD
Written by: Zachary Valley, Department of Urology, University of California-Irvine at the 2018 AUA Annual Meeting - May 18 - 21, 2018 – San Francisco, CA USA