RPS 1007 - Imaging of renal and bladder malignancies: advances in diagnosis and characterisation

Author Block: L. Chen, L. Mao, X. Li, X. Zhang, X. Bai, G. Zhang, H. Xue, Z. Jin, H. Sun; Beijing/CN
Purpose: This study developed a deep learning model for bladder and tumor segmentation from CT Urography images (CTU), integral to a system aiding bladder cancer management.
Methods or Background: A 381-case dataset from two centers, approved by the Institutional Review Board, was utilized. It comprised 280 training cases (275 with masses, 5 normal), 56 internal validation cases (54 with masses, 2 normal), and 45 external validation cases (44 with masses, 1 normal). The set included nephrographic phase CTU scans with 0.625mm and 1mm slice thicknesses for patients with pathologically confirmed lesions. A reference standard with manual contours was provided by an experienced radiologist and reviewed by a senior one. The nnU-Net framework trained a U-Net-based segmentation model using an ensemble prediction in a five-fold cross-validation and test-time augmentation. Performance was assessed using DSC, 95% HD, and MSD on the testing set.
Results or Findings: Results show that our approach achieves superb segmentation accuracy. In the internal validation set, the U-Net-based model showed strong performance with a DSC of 97.9%, 95% HD of 0.48mm, and MSD of 3.43mm for bladder segmentation. It excelled in the external set with a DSC of 98.4%, 95% HD of 0.34mm, and MSD of 2.52mm. For tumor segmentation, the internal set results were a DSC of 76.6%, 95% HD of 3.70mm, and MSD of 19.15mm, while the external set showed a slight decrease to a DSC of 74.4%, 95% HD of 3.73mm, and MSD of 22.08mm.
Conclusion: Though the tumor segmentation was less precise than bladder segmentation, the U-Net-based model still provided satisfactory accuracy for both, excelling in bladder delineation. This model proves valuable for detecting bladder cancer and evaluating treatment efficacy.
Limitations: The limited sample size of the external validation cohort limited the generalizability.
Funding for this study: This work was supported by National High-Level Hospital Clinical Research Funding(2022-PUMCH-A-035), National High-Level Hospital Clinical Research Funding(2022-PUMCH-B-069), National High-Level Hospital Clinical Research Funding(2022-PUMCH-A-033), the Natural Science Foundation of China (Grant No.81901742), the Beijing natural Science Foundation (Grant No. L232133), and the CAMS Innovation Fund for Medical Sciences (2022-12M-C&T-B-019).
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The Pecking Union College Hospital Institutional Review Board approval was obtained (ethical approval number: I-22PJ887).

Author Block: Y. Mytsyk¹, P. Kowal², Y. Kobilnyk³, I. Dutka¹, I. Komnatska¹, A. Górecki⁴; ¹Lviv/UA, ²Wroclaw/PL, ³Przemyśl/PL, ⁴Przeworsk/PL
Purpose: The aim of this study was to evaluate the effectiveness of MRI tumor parameters in predicting tissue expression of miR-15a in renal cell carcinoma (RCC) patients.
Methods or Background: The study involved 64 patients with histologically confirmed conventional RCC, where miR-15a expression was measured, and preoperative contrast-enhanced MRI (1.5 T) was performed. MiR-15a expression was determined using reverse transcription and real-time PCR. A polynomial regression model assessed associations between miR-15a expression and radiological tumor parameters, with accuracy evaluated by the Fisher method (adjusted R²).
Results or Findings: It was found that radiological features of the cystic component, exophytic growth, necrosis, macroscopic fat, and nodular contrast enhancement of the tumor were observed in 29.69%, 23.44%, 32.81%, 20.31%, and 37.5% of patients, respectively. The mean levels of miR-15a expression in the presence of these features were 0.35±1.02 U, 0.34±1.09 U, 4.01±3.42 U, 0.29±0.87 U, and 2.91±3.24 U, respectively. In the absence of these features, the mean expression of miR-15a was 2.01±2.93 U, 1.88±2.85 U, 0.82±1.85 U, 1.83±2.83 U, and 0.68±1.72 U, respectively (p<0.05). The highest miR-15a expression levels were observed with necrosis, and the lowest with macroscopic fat (p<0.05). Tumor size strongly correlated with miR-15a expression (r=0.724; p<0.001). Tumor size alone predicted miR-15a expression with an adjusted R² of 0.8281, and combining tumor size with other radiological features predicted 85% of miR-15a expression (R²=0.8336; p<0.001). The study developed a predictive formula for miR-15a expression based on RCC radiological features.
Conclusion: The findings suggest that MRI parameters can accurately predict miR-15a expression, which holds diagnostic and prognostic value in RCC.
Limitations: The main limitation was the inclusion of only conventional RCC.
Funding for this study: No funding.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The study was approved by the Ethics Committee of the Danylo Halytsky Lviv National Medical University (protocol No. 5 dated May 25, 2021). The work was conducted in accordance with accepted standards for conducting research in the field of biology and medicine, based on the guidelines of the World Health Organization, the International Council of Medical Scientific Societies, the International Code of Medical Ethics (1983), the Helsinki Declaration adopted by the General Assembly of the World Medical Association, the Convention on Human Rights and Biomedicine of the Council of Europe (1997), and the requirements and standards of ICH GCP (2002). In each specific case, patients or responsible individuals provided written consent for the surgical intervention.

Author Block: J. Han¹, T. Liu², J. Li¹, Y. Zhang³; ¹Hohhot/CN, ²Guangzhou/CN, ³Zhuhai/CN
Purpose: To determine whether the artificial intelligence integrated model based on automatic segmentation of CT images can provide a robust prediction of clear cell renal cell carcinoma (ccRCC) ISUP/WHO grade.
Methods or Background: Pretreatment CT scans were retrospectively acquired in patients with surgically proven ccRCC at multiple centers from January 2017 to September 2023.The proposed framework comprised five modules, including a 3D tumor segmentation model by 3D-UNet, a deep learning feature extraction module, a radiomic feature extraction module, a clinical-radiological feature screening module, and a fully-connected classification module that combines features from different sources to classify low-grade (I and II) and high-grade (III and IV) ccRCC. The Grad-CAM method and SHAP method are used to analyze the interpretability of the artificial intelligence model.
Results or Findings: The training data set was comprised of 335 patients from three centers, and 110 and 84 patients were included in the two external test data sets. The average Dice coefficient of the 3D-UNet automatic segmentation network in the test sets was 0.86 and 0.82. Synchronous distant metastasis, Planned nephrectomy type, and tumor long axis as independent predictors of high-grade ccRCC. In the test sets, the AUC and accuracy of integrated model were 0.85-0.92, 78-85%, respectively, which were exceeded those of clinical-radiological feature model (0.85 vs0.75 [P = 0.039], 0.93 vs 0.76 [P = 0.043], 78% vs 65% [P < 0.001], 85% vs 71% [P = 0.035].
Conclusion: An integrated model based on clinical features, radiological features, radiomics features, and deep learning features provided reliable prediction of WHO/ISUP grade for ccRCC, which outperformed the clinical-radiological feature model.
Limitations: This is a retrospective study and we only included ccRCC with a pathological diagnosis of WHO/ISUP grading after nephrectomy , with some selection bias.
Funding for this study: This work was supported in part by the National Natural Science Foundation of China under Grant Nos. 81801809, 82371917, 81830052, 81971691, 12126610, 62371476; in part by the Basic and Applied Basic Research Foundation of Guangdong Province under Grant Nos. 2020A1515010572, and in part by the Zhuhai Basic and Applied Basic Research Foundation under Grant Nos. ZH22017003200001PWC.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: This study was approved by the institutional review boards of the Fifth Hospital of Sun Yat- sen University.

Author Block: L. Melges, C. Torres, F. Chahud, L. Colli, J. Elias, C. A. Fernandes Molina, M. Castro, V. F. Muglia, D. F. Maia; Ribeirao Preto/BR
Purpose: Adrenal vein involvement is a common feature of adrenocortical carcinomas, but its prevalence in metastatic adrenal lesions remains unknown. Our goal was to assess the prevalence of adrenal vein involvement in primary and metastatic adrenal lesions and to determine if morphological changes in tumor shape precede venous extension.
Methods or Background: This retrospective, single-center observational study evaluated 102 patients: 28 adrenal cortical carcinoma (ACC) patients; and 74 non-ACC cancer patients that presented adrenal metastasis (82 metastatic adrenal lesions). Two readers reviewed cross-sectional imaging to assess tumor size, laterality, venous invasion, and the presence of the "edge sign." Surgical and histopathological confirmation was the reference standard for ACCs, while for metastases, sequential imaging or PET-CT results showing hypermetabolism were used when histopathology was unavailable.
Results or Findings: Of the 28 ACC patients, 82.1% were female, with balanced laterality. Metastases primarily originated from the lung (24.4%), colorectal (13.4%), and breast (12.2%) cancers and had a left-side dominance (61.7%). Venous extension was present in 14.6% of metastases and 21.4% of ACCs. The "edge sign" was more frequently observed in metastatic lesions (26.8%) than in ACCs (17.8%). Interobserver agreement was almost perfect for venous extension (κ = 0.9256) and substantial for the edge sign (κ = 0.7844).
Conclusion: Venous extension was less prevalent in metastatic adrenal lesions compared to ACCs. The edge sign may precede venous extension, especially in metastatic cases. Although these findings could impact clinical evaluation, prospective multicenter studies are needed to confirm the clinical significance of the edge sign.
Limitations: Retrospective, single-center study.
Funding for this study: FAEPA - Foundation for the development of learning, assistance and research of Ribeirao Preto School of Medicine Hospital.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The study was approved by the Ethics and Research Committee of the Clinical Hospital at Ribeirao Preto Medical School under number CAAE 78221024.2.0000.5440, with a waiver from the Informed Consent Form

Author Block: E. Jambon¹, N. Grenier¹, C. Marcelin¹, A. Crombé¹, G. Margue¹, J-C. Bernhard¹, F. H. Cornelis²; ¹Bordeaux/FR, ²New York, NY/US
Purpose: Small renal tumors are increasingly being detected incidentally, posing diagnostic challenges. Up to 23% of small renal tumors result in non-contributive biopsies. This study evaluates the diagnostic value of multiparametric MRI (mpMRI) in the characterization of small solid renal tumors.
The aim was to assess the diagnostic accuracy of mpMRI in differentiating malignant from benign small solid renal tumors in patients with suspected malignancy but no evident signs of metastasis.
Methods or Background: This is a prospective multicentric French study. A cohort of 387 patients in 17 centers with non-hereditary, solid renal masses between 1.5 and 4 cm in diameter was enrolled between November 2018 and May 2022. MRI protocols included T1w, T2w, diffusion-weighted imaging, and dynamic contrast-enhanced sequences. Radiologists performed blinded readings with a centralized review in case of discordance. The primary endpoint is the negative predictive value (NPV) of a dichotomized Likert scale score, targeting a 98% NPV.
Results or Findings: The study found a 45% NPV for mpMRI, falling short of the expected 98% due to difficulties in distinguishing clear cell renal cell carcinoma (ccRCC) from oncocytomas, which constituted 80% of the benign tumors in the cohort. Despite this, mpMRI influenced clinical management decisions, increasing "surveillance without biopsy" by 25% and reducing biopsies by 42%. However, it also led to a 25% increase in partial/total nephrectomies.
Conclusion: While mpMRI showed limited ability to accurately distinguish certain benign lesions from malignant tumors under predefined criteria, it proved effective in identifying malignant cases. This led to a shift in clinical management favoring surgical interventions over biopsies. The study highlights the need for more objective imaging criteria and further research into quantitative measures and radiomic analysis for better tumor characterization.
Limitations: Predefined criteria
Funding for this study: No funding
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: PHRC-K
Approval from the French Ethics Committee (CPP)

Author Block: M. N. Tasdemir, U. Eryürük, S. Aslan; Giresun/TR
Purpose: To assess the diagnostic performance of a modified biparametric VIRADS (mbp-VIRADS), derived from a combination of ADC measurements and biparametric MRI (bp-MRI), in predicting detrusor muscle invasion in bladder cancer (BC).
Methods or Background: Patients with histopathologically confirmed BC between June 2020 and May 2024 were analyzed retrospectively. Two image sets, biparametric MRI (set 1) and multiparametric (mp) MRI (set 2), were formed. Tumors were categorized using both the bp-VIRADS and mp-VIRADS systems. The optimal ADC value to differentiate muscle-invasive bladder cancer (MIBC) from non-muscle-invasive bladder cancer (NMIBC) was determined using a receiver operating characteristic (ROC) curve. To assess the mbp-VIRADS category, for cases with a bp-VIRADS score of 2-4, scores were upgraded for those below the ADC cut-off value and downgraded for those above it.
Results or Findings: A total of 182 patients with BC met the study criteria. Of these patients, 146 had NMIBC and 36 had MIBC. Comparing VIRADS categories with MIBC detection, AUC of the ROC analysis was 0.896, 0.940, and 0.941 for the bp-MRI, mbp-MRI, and mp-MRI protocols, respectively. The sensitivity for bp-VIRADS, mbp-VIRADS, and mp-VIRADS scores (with a cutoff ≥4) were 78%, 88%, and 90%, specificity were 91%, 95%, and 93%; and overall accuracy were 88%, 92%, and 93%, respectively. Using bp-MRI, there were 12 false-positive and 9 false-negative cases for predicting muscle invasion. With mp-MRI, false positives decreased to 9, and false negatives to 4. When using mbp-MRI, false positives further decreased to 6, with 5 false negatives.
Conclusion: By combining ADC measurements with bp- MRI features, the diagnostic performance of bp-MRI in predicting muscle invasion of bladder cancer can be significantly improved, approaching that of mp- MRI, while reducing the false-positive and false-negative rates.
Limitations: This was a retrospective study,
Funding for this study: None
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The study was approved by the local ethics commitee.

Author Block: A. Dehghanpour, M. Pecoraro, L. Laschena, M. Bicchetti, C. Catalano, V. Panebianco; Rome/IT
Purpose: The aim of this study was to assess the diagnostic accuracy of the VI-RADS score and its individual MRI categories, structural category (T2W), diffusion category (DWI), and contrast-enhanced category (DCE), in patients who underwent diagnostic transurethral resection of bladder tumor (TURBT). Additionally, we correlated the diagnostic accuracy of VI-RADS and its individual sequences with the time interval between TURBT and MRI, to suggest the optimal timing for MRI after TURBT. We also calculated the inter-reader agreement in scoring VI-RADS after TURBT.
Methods or Background: This retrospective single-center study included 150 patients who underwent mpMRI after TURBT at varying intervals. Four experienced readers in bladder MRI, independently and blinded to clinicopathological information, evaluated the scans, providing both VI-RADS scores and local staging. Each evaluation was performed twice: once with DWI as the dominant sequence and once with DCE as the dominant sequence.
The only exclusion criterion was prior systemic therapy. Histopathological results from therapeutic TURBT or radical cystectomy were used as the reference standard.
Results or Findings: The AUC for VI-RADS in detecting muscle-invasive bladder cancer was 0.88 (95% CI 0.84-0.92) for the most experienced reader. On a per-sequence analysis, DWI showed the highest AUC (0.83 [95% CI 0.78-0.87]), followed by DCE (0.68 [95% CI 0.63-0.74]). The diagnostic accuracy of VI-RADS improved when the time between TURBT and MRI exceeded 2 weeks and became optimal after 4 weeks, regardless of whether DWI or DCE was the dominant sequence.
Conclusion: The earliest acceptable timing for MRI after TURBT is at least 2 weeks, with the optimal timing being after 4 weeks.
In scoring VI-RADS after TURBT, DWI should be considered the dominant sequence, due to its high sensitivity and specificity.
Limitations: Retrospective design and readers being from the same center.
Funding for this study: None.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Institutional Ethical Committee.

Author Block: E. Yuan¹, Q. Zhou², Y. Chen¹, C. He¹, L. Ye¹, J. Yao¹, B. Song¹; ¹Chengdu/CN, ²Shanghai/CN
Purpose: To develop and validate a deep learning system for predicting pathological T3a upstaging in renal cell carcinoma while integrating prediction uncertainty to facilitate more reliable clinical decision-making.
Methods or Background: In this retrospective study, we collected pathology-confirmed RCC patients who underwent surgical resection from three tertiary academic medical centers. The data in first center were split into training and testing datasets. Three DenseNet-121 models were trained to predict the overall T3a invasion, the inner invasion, and the outer invasion. The uncertainty was quantified by ensemble-based and voting-based methods. For uncertain cases, manual interpretation was performed to obtain the final prediction. The performance of the pure model and uncertainty-aware interactive system were evaluated and compared on the testing dataset and two external datasets using area under the ROC curve (AUC).
Results or Findings: The data of 1329 patients (975:235:119) were collected and analyzed. The DL system performed worse in the uncertain group compared to the certain group of testing dataset (AUC 0.73 (95% CI: 0.61, 0.85) vs 0.81 (95% CI: 0.71, 0.90)), the external_1 dataset (AUC 0.50 (95% CI: 0.33, 0.68) vs 0.95 (95% CI: 0.90, 0.99)), and the external_2 dataset (AUC 0.82 (95% CI: 0.62, 1.00) vs 0.94 (95% CI: 0.89, 0.99)). The net reclassification index for the DL system were 0.11, 0.18, and 0.13 in testing, external_1, and external_2 datasets.
Conclusion: The uncertainty-aware interactive deep learning system effectively predicts pathological T3a upstaging in renal cell carcinoma, with manual interpretation improving performance in uncertain cases. This approach enhances diagnostic reliability, demonstrating potential for improved clinical decision-making across multiple datasets.
Limitations: The sample size in external validation datasets were limited.
Funding for this study: None
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The written consent was waived for the retrospective design.

Author Block: M. Cosenza, G. Brembilla, G. Imperiale, A. Larcher, U. Capitanio, F. Montorsi, F. De Cobelli; Milan/IT
Purpose: This study aimed to evaluate the improvement in diagnostic performance of radiologists in identifying clear cell and papillary renal tumors using the Clear Cell Likelihood Score (CCLS) before and after a structured training program.
Methods or Background: This monocentric study analyzed 60 MRI scans, including 28 cases of clear cell carcinoma and 16 cases of papillary carcinoma, all confirmed by histopathological examination. Five radiologists evaluated the scans twice: the first assessment was based solely on their prior knowledge, using a cutoff of ≥4 for likelihood of clear cell carcinoma; the second assessment followed a training session that introduced the CCLS, where a score of ≥4 was assigned for clear cell tumors and ≤1 for papillary tumors. A washout period of 4 weeks was implemented between assessments.
Results or Findings: Clear cell carcinoma overall sensitivity improved from 56% (95%CI: 47-64) pre-training to 86% (95%CI: 79-91) post-training, with a corresponding AUC for the ROC curve enhancing from 0.81 (95%CI: 0.76-0.86) to 0.85 (95%CI:0.81-0.90). Papillary carcinoma sensitivity increased from 52% (95%CI:41-64) pre-training to 65% (95% CI: 54-75) post-training, with specificity rising from 90% (95%CI:85-93) to 95% (95%CI:91-97). The AUC for the ROC curve for papillary carcinoma rose significantly from 0.79 (95%CI:73-85) to 0.89 (95%CI: 85-94). Additionally, the agreement improved for clear cell tumors, with a K of Conger increasing from 0.293 (95%CI: 0.181-0.406) to 0.594 (95%CI:0.469-0.718), and for papillary tumors, from 0.360 (95%CI:0.211-0.51) to 0.489 (95%CI:0.35-0.628).
Conclusion: Structured training and the application of the CCLS significantly enhance the diagnostic accuracy of radiologists in identifying clear cell and papillary renal tumors on MRI, underscoring the importance of targeted education in improving radiological interpretations.
Limitations: Limitations include monocentric design, small sample size, limited radiologist cohort, and no longitudinal follow-up, impacting generalizability and sustainability.
Funding for this study: None
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: IRB approved

Author Block: P. Kowal¹, Y. Mytsyk², K. Ratajczyk¹, W. Bursiewicz¹, M. Trzciniecki¹, K. Marek-Bukowiec¹, J. Rogala¹; ¹Wrocław/PL, ²Lviv/UA
Purpose: This study aimed to evaluate the utility of first-order radiomic features derived from MRI apparent diffusion coefficient (ADC) maps using volumetric analysis in distinguishing solid from friable thrombus in patients with renal cell carcinoma (RCC).
Methods or Background: A cohort of 27 patients with conventional histologic subtype of RCC and tumor thrombus in the renal vein or inferior vena cava (IVC) was included. All patients underwent surgical intervention, comprising nephrectomy and thrombectomy, and received preoperative abdominal MRI with diffusion-weighted imaging sequences at b-values of 50, 200, 800 s/mm². The ADC map was used for volumetric analysis, calculating various radiomic first-order features across the thrombus volume, including ADC mean, median, range, 10th percentile, 90th percentile, interquartile range, entropy, kurtosis, skewness, uniformity, and variance. Tumor thrombi were histologically classified as solid or friable, and associations between the radiomic features and thrombus consistency were analyzed.
Results or Findings: Solid and friable tumor thrombi were identified in 51.9% and 48.1% of patients, respectively. Inverse association noted between RCC thrombus cellularity and skewness (r=-0.799, p<0.001). No significant differences were observed in the mean values of range, 90th percentile, interquartile range, kurtosis, uniformity, and variance between groups. For distinguishing solid from friable thrombus, the ADC mean, median, and entropy showed equal sensitivity (93%) and specificity (69%), with entropy yielding the highest area under the curve (AUC) at 0.808. Skewness demonstrated a sensitivity of 86% and specificity of 92%, with an AUC of 0.931.
Conclusion: In RCC patients with tumor thrombus in the renal vein or IVC, volumetric analysis of first-order radiomic features using ADC mapping facilitates accurate differentiation between solid and friable thrombus variants.
Limitations: The primary limitation of this study is that only conventional histologic subtype of RCC was included in the analysis.
Funding for this study: No funding.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: This study was approved by the Local Bioethical Committee in the Research and Development Center, Regional Specialist Hospital in Wroclaw (no. KB/12/2021). All procedures conducted followed the ethical guidelines set by the institutional and/or national research committee, adhering to the principles outlined in the 1964 Helsinki Declaration and its subsequent revisions, or equivalent ethical standards. The written informed consent for enrolment in the study was signed by all patients.

Author Block: Ş. Evrimler, E. Cigdem Karatayli; Ankara/TR
Purpose: The Bosniak Classification system was revised in 2019 to reduce subjectivity and prevent unnecessary nephron loss. In this plot study, we aimed to evaluate the inter-reader agreement in Bosniak classification among radiologists with different experience levels.
Methods or Background: Out of 320 patients imaged between 01.2022 and 04.2024 in our hospital, 12 patients were randomly selected among those with pathology results, 22 patients were randomly selected among those without pathology results, a total of 34 patients were selected. Eight residents, two radiology specialists, and one abdominal radiologist evaluated abdominal CT and MRI scans of these patients. All participants were trained on the 2019 Bosniak classification prior to the assessment. They assessed each criteria (septation presence, nodularity characteristics, contrast enhancement,etc.) separately and ultimately identified the Bosniak type. The abdominal radiologist's classifications served as the reference for Kappa analysis. Participants also completed a questionnaire regarding difficulties in evaluating the classification and their use of objective values.
Results or Findings: Agreement with the reference increased with radiology training duration (min kappa: 0.47, max kappa: 0.79, p<0.01). Substantial agreement was noted between specialists and the reference (kappa: 0.68-0.72, p<0.01). Agreement was better for non-measurable parameters (e.g., presence of septation) compared to measurable ones (e.g., thickness of septation). Survey results indicated the most challenging parameter was septation thickness, with 100% of participants uncertain between categories 2 vs. 2F and 2F vs. 3. While 53% used objective measurements during assessments, 81% relied on them as a guide rather than exclusively.
Conclusion: Moderate-substantial agreement was found among radiologists, improving with experience. Thickness of septations was particularly confusing, especially between certain categories.
Limitations: The study's small sample size suggests further research with larger, diverse groups is needed.
Funding for this study: No funding was provided for this study.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: AEŞH-EK1-2024-54 numbered file, Ethical Committee, Etlik City Hospital, Ankara