Ethical aspects

This work was approved by the Ethics Committee on the Use of Animals at the University of Franca (n. 4748070318).

Selection of animals

The experiment was performed at the Veterinary Clinic Pronto Socorro Veterinário in Uberlândia-MG. In this study, six dogs cadavers, three females and three males were obtained from hospital care, after due release and authorization by their owners. The average weight was 7.7 ± 3.9 kg and the average age of 7.6 ± 5.2 years. Cadavers that had no abnormalities on visualization and interpretation of the bladder and urethra on x-ray were selected. As an exclusion criterion, all selected animals had previously been evaluated by survey radiograph and were free from any alteration in the anatomical conformity of the lower urinary tract, as well as overlapping artifacts like gases and feces. The animals were kept frozen until the moment of the experiment. These animals were thawed naturally in an air-conditioned environment.

Implantation of uroliths

One of the cadavers was randomly selected as a control (no uroliths were inserted) and in the others, ammonium urate uroliths were implanted in the urinary bladder using a urethral catheter. These uroliths were obtained from clinical routine and classified according to mineral type by the Nephrology Service of the Veterinary Hospital of the University of Franca (UNIFRAN).

The selection of the number of implanted uroliths was carried out randomly, with numbers from 1 to 30 (based on our hospital routine—data not shown) being recommended for the draw. All uroliths used had sizes ranging from 1 to 3 mm, with 3, 7, 8, 12, and 15 randomly selected uroliths being inserted into the animals. The Random® application (“Aleatório”, in portuguese) version 1.2 (https://play.google.com/store/apps/details?id=com.start.devaneio.alatorio) was used to select the number of uroliths by the study supervisor, who did not perform stone counts during the examinations. In females, an episiotomy was performed to introduce the uroliths via pelvic urethra with the aid of a urethral catheter n°8—Mark Med. In males, the perineal urethra was exposed and sectioned to secure the urinary catheter n°8 –Mark Med for the introduction of uroliths, all catheters were fixed by a suture in tobacco pouch to avoid extravasation of the applied contrast. There was no accidental removal of any uroliths during contrast drainage or bladder massage.

Radiographic projections

All animals were submitted to two radiographic projections with simple x-ray, in the lateral and ventro-dorsal positions; After this initial step, all animals were submitted to positive and double-contrast cystography radiographs (lateral and ventro-dorsal) with different contrast concentrations.

Positive cystography

Positive cystography techniques were performed in ascending order of contrast concentration. The contrast used in all exams was Iopamidol 612 mg/ml (Iopamiron® 300, Patheon Italy S.p.a). Contrast concentrations of 5%, 10% and 20% were used, adapted from the original iodinated contrast (Omnipaque 300mg I/mL) Another two techniques (30% and 50% concentration) were performed in one animal only, due to the impossibility of diagnosing the uroliths.

The concentrations of positive cystography mentioned above were obtained from the pure contrast diluted in saline solution, and the dose of contrast applied was 10 ml/kg [6, 7]. Between each change of concentration, the bladder was emptied with the help of a 20 ml syringe coupled to the urethral catheter, the volume inserted was removed in its entirety for the beginning of the next concentration to be evaluated.

Double-contrast cystography

Three distinct double-contrast cystography standards were performed, following established protocols. The first one (protocol 1) began with an intravesical application of 4.5 ml of pure contrast (Iopamiron® 300) per animal, followed by the application of 10 ml/kg of atmospheric air, without removing the contrast [7]. Protocol 2 started with 5 ml and 10 ml of pure contrast applied to animals weighing up to 10 kg and above 10 kg, respectively. After injecting the pure contrast, a light bladder massage was performed and then the excess contrast was removed with syringe suction. Afterward, the air was applied at a dose of 8 ml/kg. Finally, protocol 3 tested 10 ml of pure Iopamiron® 300 intravesical contrast (literature dose 5–15 ml/animal), associated with the application of 9 ml/kg of air (literature dose 6–12 ml/kg of air), without removing the contrast [7]. During the changing of contrasts of different concentrations, the urinary bladder was emptied completely, and the application of the next concentration began. At the end of the procedure, the bladder was emptied and the inserted uroliths were fully recovered.

Radiological images and assessments

The equipment used to obtain digital images was a CDK x-ray device, model MAG VET, 320 mA and 125 kV. In total, 86 digital radiographic images were obtained. After, they were submitted to three “blind” evaluators who were veterinarians specialized in veterinary radiology with an evaluation spreadsheet.

All digital radiographic images (86 radiographic images) were obtained in DICOM [8]. Then, random codes were created containing five digits (randomly between letters and numbers)–“Random” application, version 1.2. The number of codes created was proportional to the number of radiographic images obtained, and each code was used to rename each digital image, so the images were then scrambled for later evaluation.

Access to the digital images in DICOM was available to the evaluators, which were evaluated by the number of uroliths diagnosed in the bladder (values from 0 to 30), amount of contrast (insufficient, adequate, or excessive), quality of images provided (adequate, artifacts). To be considered adequate and without artifacts, the image had to be categorized in this way by the three evaluators.

Statistical analysis

For statistical analysis, the MedCalc® program for Windows in version 18.11.3 was used (MedCalc Software, Ostend, Belgium). Sensitivity and specificity tests were performed for each positive cystography evaluation (5%, 10%, and 20%), as well as for double-contrast protocols. As a criterion of truly positive values, 10% standard deviation was used under the total number of uroliths inserted. False-positive values were found when the number of uroliths was higher than those inserted, or if the negative control animal was categorized with the presence of uroliths. The truly negative was categorized in the negative control animal, which did not have uroliths inserted and was recognized as negative by the evaluators. Animals with the presence of uroliths, which were classified as exempt by the evaluators, these were identified as false negatives.

For the evaluation of the results among the three evaluators, the Kappa statistical method was used to assess the degree of agreement between the classes evaluated using the intraclass correlation coefficient.

This method uses a descriptive scale for values, where ρ is Pearson’s correlation coefficient, and ρ c represents the correlation coefficient, being: ρ c < 0.90 (weak agreement force); 0.90 < ρ c < 0.95 (moderate agreement strength); 0.95 < ρ c < 0.99 (substantial agreement strength); and ρ c > 0.99 (almost perfect agreement strength), and a 95% confidence interval [13–15].