The Computed Tomography (CT) images of the candidate for training module were used for 3D printing after receiving the written informed consent. No identifier was shown in the trainee, and all experimental protocols were approved and consent for volunteers was exempted by the institutional review board of the Seoul National University Hospital Biomedical Research Institute (IRB No. E-2106-023-1224) because only volunteers joined this education program among 32 residents and 14 board-certified urologists in the urology department of the authors’ hospital. Therefore, only twelve residents and five young board-certified urologists with no experience of RIRS as operators joined this study. All methods were carried out in accordance with relevant guidelines and regulations.
Simulator production using 3D printing
As shown in Fig. 1, the 3D printing simulator was manufactured through five processes as described below:
(1) Segmenting
The basement of simulator’s kidney anatomy was extracted by segmenting CT Digital Imaging and Communications in Medicine (DICOM) images using MEDIP (Medical IP, Seoul, Republic of Korea), a reconstruction and rendering program for segmentation, with STereoLithograph (STL) file extension (Fig. 1A).
Structures of renal major and minor calyces, renal pelvis, and ureters were primarily rendered through segmentation and transformed into a form suitable for training based on anatomical information. The authors in this article confirmed kidney structures several times through ‘Visual printing’. This is an efficient real-time confirmation platform between urologists and Medical IP. Once a web address of a rendered object was uploaded to the server, it was sent to the authors (Fig. 1B). The authors could then annotate directly on the web page via text or drawing to give feedback using a smartphone or a computer.
(2) 3D Mesh modeling
After the anatomy was confirmed, an acrylic plate was designed equipped with a tubing line and a water tank for application of a water circulation system (Fig. 1C). Classification and Boolean operation were performed using a modeling program for no overlapping part of the designed STL file. We then extracted each STL file and uploaded it to GrabCAD (Stratasys 3D Printer Slicing program) (Fig. 1D).
For each STL file of each structure transferred to GrabCAD, the desired material and color transparency were independently selected. Then a transparent clear resin was used to the exterior so that the inside structure could be seen with naked eyes from the outside. In the case of calyces and pelvis, a pink translucent material (Vero clear plus Vero magenta v mixture) was used (Fig. 1E).
(3) 3D printing
After selecting all materials, 3D Printing was performed using a Polyjet printer J750 (Stratasys; 7665 commerce way Eden prairie, MN 55344, USA) as shown in Fig. 1F.
(4) Post-processing
The supporter on the printed material was removed. Locking rings were attached onto both sides to fasten anterior and posterior sides. The sling ring was then fixed with silicon on the bottom to prevent leakage while water circulated. A mold of sling ring was printed out of Acrylonitrile Butadiene Styrene copolymers using FlashForge, a Fused Deposition Modeling type printer (Fig. 1G). After that, fumigation was performed to remove the lattice-shaped hard support and smoothen the surface. After assembling the mold according to the position, a silicon of Shore A10 was injected and molded.
(5) Assembly
The completed module was fixed to a custom-made acrylic plate and connected to the tube line. The whole procedure was finished by connecting a water pump to the end of the inlet tube and installing it in a water tank.
(6) Design and the structure of the simulator
Because the module had its own water circulation and water-proof systems, it only needed an endoscopy system for RIRS training. The tank and kidneys were filled with water and circulated with an electronic pump to create similar environment to the real practice of RIRS. LithoVue™ (Boston Scientific, Marlborough, MA, USA) was used for experiments to reduce total volume of the training set as shown in Fig. 1H. The kidneys can be split into two anterior and posterior pieces and stone fragments can be put inside the calices. The training of stone fragmentation and basketing technique is available.
Instructions and orientation education
Two expert endourologists provided the orientation training over a 10-minute session to all trainees. This included the information as follows: (1) an explanation of the caliceal anatomy, locations of upper, mid, lower calyces, and their anterior and posterior branches of minor calyces; (2) six movement directions of the fURS including upward, downward, left-sided, right-sided, forward, and backward; (3) skills of thumbs’ up and down motion and flexion and extension of the wrists with the help of left fingers are shown in Figs. 2(A) and 2(B); (4) differences in motion between right and left-sided kidneys; (5) ‘the reference point’, where the navigation starts, defined as the point when the fURS is located at the level of ureteropelvic junction where the scope can see the inlets of upper pole in front and mid calyces at lower screen at a time as shown in Fig. 2(C); (6) the direction from the upper to lower poles, it normally lies from 1 to 7 o’clock in the right-sided kidney and from 11 to 5 o’clock in the left-sided kidney; (7) the location of the inlet of the lower pole. When fURS is flexed, the inlet to the lower calyces can be identified. When the orientation was lost, the authors usually recommended to return to this reference point and then start navigation again.
RIRS navigating training protocol
The 3D printing model consisted of three minor calyces in upper and lower poles and two in the mid calyx (i.e., eight calyces in a single kidney). Each calyx was marked with a red dot on the anterior side and a blue dot on the posterior side. Thus, there were a total of 16 points per kidney. Because upper pole calyces are located most dependently, the navigation should be performed from the renal pelvis followed by the lower anterior, lower posterior, mid anterior, mid posterior, upper anterior, and upper posterior calyx in order. Fragmented stones normally gather in the upper posterior and mid posterior calyces. The final dusting procedures can be performed in these two calyces efficiently.
Evaluation parameters
The navigation time was measured as the primary outcome, and the satisfaction of the trainee was surveyed as the secondary outcome.
One navigation time started from the moment when the endoscope entered into the ureter of the model. All 16 points were checked when the scope approached the inner side of calyces where laser emission and stone fragmentation were available in clinical situations. After navigation was completed, the navigation time was measured and recorded. Basically, the goal was to perform ten times of navigation for each side, starting from the right or the left side at random. Once trainees began from the right side, they performed it on the left side the next time (RL group). If trainees began from the left side, they performed it on the right side the next time (LR group).
The experiment was stopped when specialists judged that the trainee reached the plateau of no more time reduction three times or more or that the trainee’s technique was sufficiently completed by performing navigation smoothly without hesitation. The training for stone fragmentation and basketing technique is not performed for this study to simplify the protocol.
After the experiment, all trainees were asked to fill out a questionnaire. The questionnaire consisted of 7 questions related to the following: (1) similarity to real kidney, (2) training efficacy to the upper calyx, (3) training efficacy to the mid calyx, (4) training efficacy to the lower calyx, (5) the usefulness of red and blue dots, (6) usefulness for RIRS practice, and (7) commendableness of tests to other beginners. Each question was evaluated in five degrees (5: strongly agree, 4: agree, 3: neutral, 2: disagree, 1: strongly disagree).
Statistical analysis
All statistical analyses were performed using IBM-SPSS 25.0. (New York, NY; United States). For categorical variables, results are expressed as numbers and percentages. Chi-squared test and Fisher's exact test were used for categorical variables. Continuous variables are expressed as median values. ANOVA and Student t-test were used to compare variables in two or more categories. The Wilcoxon signed rank test was used to compare the first and the last trials. Statistical significance was considered when a bilateral p-value was less than 0.05.