Koray Kaya Kilic

Aims: To systematically compare the candidate-level classification performance of four state-of-the-art deep learning architectures—U-Net, ResNet-50, Vision Transformer (ViT-B/16), and YOLOv8—for pulmonary nodule candidate classification using patch-based classification on public benchmark CT datasets, and to evaluate the trade-off between detection accuracy and computational efficiency.Methods: This computational study utilized publicly available, de-identified datasets including LUNA16 (Lung Nodule Analysis 2016) and LIDC-IDRI (Lung Image Database Consortium). Candidate nodule locations were generated using a multi-scale Laplacian of Gaussian (LoG) blob detector applied to full CT volumes. From these candidates, 64×64×64 voxel patches were extracted and classified as true nodules or false positives. The dataset was partitioned at the patient level: 70% training, 15% validation, and 15% held-out test. Stratified 5-fold cross-validation was conducted exclusively within the training set for hyperparameter optimization. Four deep learning architectures were trained under identical protocols: U-Net (encoder decoder), ResNet-50 (residual CNN), ViT-B/16 (self-attention transformer, adapted to 3D patch input via 3D patch embedding), and YOLOv8 (real-time detector, applied slice-by-slice with 3D aggregation). Primary performance metrics included sensitivity, specificity, F1-score, mAP@0.5, and area under the ROC curve (AUC). Free-response ROC (FROC) analysis was performed following LUNA16 challenge standards, reporting sensitivity at 0.125, 0.25, 0.5, 1, 2, 4, and 8 false positives per scan (FP/scan). Statistical comparisons focused on AUC using paired DeLong’s test with Bonferroni correction for multiple comparisons. Bootstrap confidence intervals (n=2,000 resamples) were computed for sensitivity, specificity, and F1-score. Results: Across 888 CT scans (1,186 annotated nodules; LUNA16 test set: 133 scans, 178 nodules), Vision Transformer achieved the highest candidate-level patch classification performance: sensitivity 94.2% (95% CI: 91.8–96.1%), specificity 92.8% (95% CI: 90.3–94.9%), F1-score 0.935, mAP@0.5 0.947, and AUC 0.971 (95% CI: 0.958–0.982). Pairwise AUC comparisons using DeLong’s test confirmed superior discrimination for ViT-B/16 relative to the comparator architectures. FROC analysis demonstrated ViT-B/16 achieved the highest mean sensitivity at 7 operating points (CPM=0.847), outperforming ResNet-50 (CPM=0.798), YOLOv8 (CPM=0.781), and U-Net (CPM=0.762). However, ViT-B/16 required 3.2× longer inference time (8.4 vs 2.6 seconds/scan) and 3.7× more trainable parameters than ResNet-50. YOLOv8 demonstrated superior computational efficiency with the shortest inference time (1.1 seconds/scan). Conclusion: The attention-based Vision Transformer architecture achieved superior candidate-level patch classification performance for pulmonary nodule candidate evaluation; however, this advantage must be weighed against substantial computational costs. Architecture selection should be guided by deployment context, with ResNet-50 offering optimal accuracy efficiency balance for clinical deployment and YOLOv8 for real-time screening applications.

Background/Objectives: Medial meniscus extrusion (MME) is a quantitative marker of altered meniscal containment and load sharing. Although ultrasonography enables dynamic assessment under functional loading, it remains unclear whether generalized ligamentous hypermobility influences physiologic extrusion behavior in healthy knees. The aim of this study was to quantify load-dependent MME in healthy adults and to determine whether generalized hypermobility is associated with greater physiologic extrusion under progressive loading conditions. Methods: In this prospective observational study, 106 healthy adults aged 18–40 years were evaluated between October and December 2025. Generalized joint hypermobility was defined as a Beighton score ≥5. MME was measured by standardized ultrasonography on the dominant limb in three conditions: sitting (unloaded), bipedal stance, and unipedal stance. Patellar tendon shear-wave elastography (SWE) was recorded in kilopascals (kPa). Interobserver reliability was assessed in the first 25 participants using ICC (2,1). Group comparisons, multivariable linear regression for loading-related Δ-extrusion (Unipedal−Sitting and Bipedal−Sitting), and a linear mixed-effects model for repeated MME measures, including a Position × Hypermobility interaction, were performed. Results: Twenty-eight participants (26.4%) were classified as hypermobile. The hypermobile group showed significantly lower patellar tendon SWE than controls (23.8 ± 7.0 vs. 37.6 ± 9.7 kPa, p < 0.001). MME increased stepwise with loading in both groups and remained consistently higher in hypermobile participants across sitting, bipedal, and unipedal conditions (all p < 0.001). Loading-related extrusion was also greater in the hypermobile group for both Bipedal−Sitting (p = 0.037) and Unipedal−Sitting (p = 0.002). In multivariable regression, lower patellar tendon SWE independently predicted greater loading-related extrusion, whereas hypermobility status did not remain an independent predictor. In the mixed model, the Position × Hypermobility interaction was significant and was most pronounced during the unipedal stance. Conclusions: In healthy adults, medial meniscus extrusion increases stepwise from unloaded sitting to bipedal and unipedal weight bearing. Participants with generalized hypermobility demonstrated higher physiologic MME values and a more pronounced load-dependent pattern, particularly during unipedal stance. However, in adjusted analyses, lower patellar tendon stiffness on SWE, rather than hypermobility status itself, independently predicted greater loading-related extrusion. These findings support a contextual interpretation of ultrasound-based MME measurements in relation to loading condition and hypermobility phenotype.

Background: Osteochondral lesions of the talus (TOLs) are a frequent cause of chronic ankle pain and functional limitation. Despite their clinical significance, data on their prevalence, anatomical distribution, and morphological features remain limited, particularly concerning sex-based differences. Methods: This retrospective study reviewed 5,356 ankle MRI scans performed over five years to identify patients with TOLs. 490 skeletally mature patients (283 females, 207 males) were included. Lesions were evaluated based on anatomical localization using a standardized nine-zone grid, graded according to the Bristol/Hepple classification, and measured using region of interest (ROI) tools. Inter- and intra-observer reliability was assessed using ICC and Cohen’s kappa statistics. Results: The overall prevalence of TOLs was 9.1%, with similar rates in males (9.3%) and females (9.0%). The mean age was 46.2±14.3 years. Lesions were most commonly located in Zones 4 (33.7%) and 7 (33.5%). The mean lesion area was 78.2±61.9 mm², with the largest lesions in Zone 4. Grade 1 (26.2%) and Grade 5 (20.70%) were the most frequent lesion grades. Male patients were significantly younger(42,2) and had larger lesions (p

Purpose: The purpose of this study was to investigate how the choice of axial CT level affects the reliability and diagnostic accuracy of the Dejour classification for trochlear dysplasia and to evaluate a novel level defined at the most superior extent of the Blumensaat line. Materials and methods: Patients who presented with patellar instability or acute patellar dislocation between 2014 and 2024 and had preoperative CT scans were retrospectively reviewed. Fifty patients were randomly selected based on an a priori sample size calculation. For each knee, four axial CT levels were reconstructed: midpatellar level, Roman arc level, 3 cm above the joint line, and the top of the Blumensaat line. A consensus Dejour grade (A–D) was established by an experienced musculoskeletal radiologist and an orthopedic sports surgeon and used as the reference standard. Two orthopedic surgeons independently graded all 200 axial images twice at least 15 days apart. Quadratic weighted kappa (κ) with 95% confidence intervals (CI) was used to assess intra- and inter-observer reliability and agreement with the consensus. Diagnostic accuracy was defined as the proportion of correctly classified cases relative to the consensus and was compared across levels using Cochran’s Q test. Results: When all four levels were combined, intra-observer reliability was almost perfect for both observers (κ = 0.96 and 0.84; exact agreement 91% and 84%), and inter-observer reliability was substantial to almost perfect (κ = 0.72 and 0.78; exact agreement 72–73%). Agreement with the consensus across all levels was moderate (κ = 0.52–0.58; exact agreement 51–52%). Analyzing levels separately, intra-observer κ remained high at all levels, whereas inter-observer agreement and agreement with the consensus varied markedly. The midpatellar level showed only moderate inter-observer reliability and fair-to-moderate agreement with the consensus (κ = 0.36; accuracy 34–40%), whereas the top of the Blumensaat line showed the highest agreement with the consensus (κ 0.69) and the highest accuracy (up to 64%; pooled 61%); however, statistically significant between-level differences were detected in only one observer–time comparison. The 3 cm above the joint line and the Roman arc level demonstrated intermediate performance. Conclusions: Although intra-observer reliability of the Dejour classification is high regardless of axial CT level, both inter-observer agreement and diagnostic accuracy depend strongly on the selected slice. The axial CT level at the top of the Blumensaat line showed a consistent trend toward higher agreement and accuracy relative to the consensus standard and may be used as a standardized reference slice within routine multi-slice CT assessment to improve reproducibility; however, it should complement comprehensive imaging review and clinical evaluation.

This study presents a quantitative MRI-based method to improve the objectivity and diagnostic accuracy of grading calf muscle strains. Ninety-six patients with MRI-confirmed medial gastrocnemius injuries who underwent bilateral lower leg MRI between January 2018 and December 2023 were retrospectively analyzed. Injuries were graded using the Dai classification (Grades I–III). Axial T2-weighted SPAIR images were used to measure the medial gastrocnemius cross-sectional area (CSA) at the distal one-third between the knee joint and the musculotendinous junction. The Gastrocnemius Medial Head Ratio (GCM ratio) was calculated by dividing the CSA of the injured side by that of the uninjured side. Two experienced radiologists independently performed the measurements twice. Intra- and interobserver reliability were assessed using intraclass correlation coefficients (ICC) and Cohen's kappa. A one-way ANOVA with the Bonferroni correction was used for group comparisons. ROC analysis evaluated the diagnostic performance of the GCM ratio. The GCM ratio significantly differed among Grade I (1.03 ± 0.02), Grade II (1.18 ± 0.06), and Grade III (1.35 ± 0.05) injuries (p < 0.001). Ipsilateral CSA increased with injury severity, whereas contralateral CSA remained consistent. CSA measurements showed excellent intra- and interobserver reliability (ICC > 0.87), and lesion grading demonstrated substantial to almost perfect agreement (κ = 0.75–0.88). ROC analysis revealed strong diagnostic accuracy (AUC = 0.986 for Grade I vs. II; AUC = 1.000 for Grade II vs. III). The GCM ratio provides a reproducible and objective MRI marker to aid in grading gastrocnemius strains, offering excellent diagnostic performance and clinical applicability.

Objectives: Hysterosalpingography (HSG) remains the first-line imaging technique for assessing fallopian tube conditions due to its accessibility, cost-effectiveness, and high diagnostic accuracy. Intravasation, the leakage of contrast media into myometrial or vascular structures during HSG, has been documented as a complication. This study aimed to determine the prevalence of intravasation, explore the relationship between intravasation and tubal occlusion, and assess the applicability of grading intravasation. Methods: A retrospective analysis of 3032 HSG examinations performed between January 2021 and May 2024 was conducted. Intravasation, tubal occlusion, and related findings were recorded using standard X-ray equipment and categorized prospectively. Statistical analysis included logistic regression to identify predictors of tubal occlusion and inter-observer agreement using Fleiss Kappa. Diagnostic metrics for intravasation as a radiological marker of tubal occlusion were calculated. Results: Intravasation was observed in 2.6% of cases. Among cases with tubal occlusion, 16.4% exhibited intravasation. The sensitivity, specificity, positive predictive value, and negative predictive value of intravasation for detecting tubal occlusion were 14.5%, 99.8%, 92.3%, and 85.6%, respectively. Adjusted odds ratios indicated that intravasation increased the likelihood of tubal occlusion approximately 55-fold. Inter-observer agreement for intravasation detection was almost perfect (Kappa = 0.97). Limitations included the absence of oil-based contrast and a pressure manometer during contrast infusion. Conclusions: Intravasation, traditionally viewed as a complication, emerges as a significant radiological indicator of tubal occlusion. Incorporating intravasation characteristics, such as initiation time, into grading systems may enhance diagnostic accuracy. Further refinement of classification systems and broader clinical validation are recommended. Advances in knowledge: About 92.3% of tubes with intravasation were occluded and the site of intravasation correlated with the occlusion site. Occlusion was 55× more likely in cases with intravasation, traditionally viewed as a complication, emerges as a significant radiological indicator of tubal occlusion.

Sarcopenia, defined as a loss of skeletal muscle mass and/or function, has been associated with poor survival in various cancers. This study evaluates the prognostic impact of sarcopenia in multiple myeloma (MM) using the Psoas Muscle Index (PMI). A total of 181 MM patients diagnosed between 2014 and 2024 were retrospectively analyzed. Sarcopenia was assessed using PMI measurements at the L3 vertebra level, with gender-specific cut-off values determined by ROC analysis (442.63 mm²/m² for males and 308.85 mm²/m² for females). Sarcopenia was identified in 29.8% of patients. Sarcopenic patients were significantly older, had lower BMI, albumin, and hemoglobin levels, and were less likely to be eligible for transplantation. Multivariate analysis revealed that older age (OR:1.14, 95%CI:1.08–1.32, p = 0.014), hypoalbuminemia (OR:2.13, 95% CI: 1.11–4.12, p = 0.023), and low BMI (OR:2.52, 95%CI:1.83–5.23, p = 0.001) were independent risk factors for sarcopenia. Sarcopenia independently predicted worse PFS (HR:1.628, 95%CI:1.065–2.487, p = 0.024) and OS (HR:2.095, 95%CI:1.207–3.636, p = 0.009). Other independent risk factors for OS included LDH > ULN (HR:2.026, 95%CI:1.130–3.631, p = 0.018), transplant eligibility (HR:0.338, 95%CI:0.191–0.596, p < 0.001), and high cytogenetic risk (HR:1.912, 95%CI:1.004–3.640, p = 0.049). Subgroup analyses showed that sarcopenia significantly affected OS in transplant-ineligible patients (HR:0.501, 95%CI: 0.262–0.957, p = 0.03) but had no significant impact on survival in transplant-eligible patients. Sarcopenia analysis by using PMI in MM patients can be considered a cost-effective, simple, easily applicable, and practical method for daily routine. Sarcopenia independently predicts worse survival outcomes, particularly in transplant-ineligible patients. Larger prospective studies are warranted to validate these findings and explore integrating sarcopenia into frailty indices for MM.

This study aimed to evaluate and compare the reliability and agreement of four commonly used methods for patellar height measurement: Insall-Salvati (IS), Modified Insall-Salvati (mIS), Caton-Deschamps (CD), and Blackburne-Peel (BP) indices, using standardized lateral knee radiographs. This prospective study included seventy adult patients with standardized lateral knee radiographs. Radiographs met strict criteria for true-lateral views and appropriate knee flexion angles. Three observers independently measured patellar height using the four methods, each performing two rounds of measurements at a 15-day interval. Intraclass Correlation Coefficients (ICC) were calculated for intra-observer and interobserver reliability. Pairwise agreements between methods were analyzed using Cohen’s kappa with 95% confidence intervals (CI). Correlation coefficients (Pearson’s r) were computed to assess the relationships between the methods. The BP method exhibited the highest inter-observer reliability (ICC = 0.848, 95% CI: 0.754–0.910) and intra-observer reliability (average ICC = 0.908, 95% CI: 0.829–0.986). This was followed closely by the CD method, while the IS and mIS methods demonstrated moderate reliability. Correlation analysis revealed strong relationships between the BP and CD methods (r = 0.871, p < 0.01) but weaker correlations between IS and the other methods. Despite high correlation coefficients, agreement in classification was poor, particularly between IS and BP (κ = -0.131) and IS and mIS (κ = -0.047).The Blackburne-Peel method is the most reliable and reproducible for patellar height measurement under standardized conditions, making it a robust tool for clinical and research applications. However, the methods were not interchangeable, emphasizing the need for consistent imaging protocols and method selection tailored to clinical objectives.

Kose O, et al. Clin Orthop Surg. 2025 Feb;17(1):100-111. https://doi.org/10.4055/cios24270

Calf muscle strain injuries (CMSI), also known as “tennis leg,” are frequently observed in middle-aged and physically active people, including non-professional athletes. While magnetic resonance imaging (MRI) is frequently used to assess the extent of these injuries, the relationship between MRI findings and long-term functional outcomes in non-athletic populations remains underexplored. This study aimed to investigate the correlation between MRI-detected injury severity and functional outcomes using the Achilles Tendon Rupture Score (ATRS) and the Tegner Activity Scale (TAS). A retrospective review was conducted on 78 non-athletic patients diagnosed with CMSI and were followed for an average of 25.6 ± 16.1 months. Injury severity was classified into three grades: Grade 1 (edema without architectural disruption), Grade 2 (partial muscle disruption with hematoma), and Grade 3 (complete muscle disruption or tendon detachment) based on MRI examination. The functional outcomes were assessed using the ATRS and TAS. A simple linear regression analysis was conducted to assess the impact of demographic variables and MRI findings on ATRS. The mean ATRS at the final follow-up was 92.9 ± 10.6, with 75.6% of patients achieving an excellent outcome. Despite the excellent ATRS scores, a notable reduction in TAS was evident from the pre-injury assessment (4.4 ± 1.2) to the final follow-up (3.8 ± 1.3; p < 0.001). Moreover, 44.9% of patients did not resume their pre-injury activity levels. The regression analysis demonstrated no statistically significant correlation between MRI-detected injury severity, demographic factors (age, gender, BMI), and ATRS outcomes (R² = 0.094, adjusted R² = -0.011). MRI-detected injury severity does not significantly predict long-term functional outcomes in patients with CMSI, even though most patients achieved excellent ATRS scores. However, nearly half of the patients did not return to their pre-injury activity levels. These findings suggest that additional factors might influence recovery, and further research is needed to elucidate these factors in non-athletic patients. Level IV, Retrospective case series.