Correction to: Deep convolutional neural network-based detection of meniscus tears: comparison with radiologists and surgery as standard of reference.

The article "Deep convolutional neural network-based detection of meniscus tears: comparison with radiologists and surgery as standard of reference.

Deep Convolutional Neural Network-Based Diagnosis of Anterior Cruciate Ligament Tears: Performance Comparison of Homogenous Versus Heterogeneous Knee MRI Cohorts With Different Pulse Sequence Protocols and 1.5-T and 3-T Magnetic Field Strengths.

OBJECTIVES: The aim of this study was to clinically validate a Deep Convolutional Neural Network (DCNN) for the detection of surgically proven anterior cruciate ligament (ACL) tears in a large patient cohort and to analyze the effect of magnetic resonance examinations from different institutions, varying protocols, and field strengths. MATERIALS AND METHODS: After ethics committee approval, this retrospective analysis of prospectively collected data was performed on 512 consecutive subjects, who underwent knee magnetic resonance imaging (MRI) in a total of 59 different institutions followed by arthroscopic knee surgery at our institution. The DCNN and 3 fellowship-trained full-time academic musculoskeletal radiologists evaluated the MRI examinations for full-thickness ACL tears independently. Surgical reports served as the reference standard. Statistics included diagnostic performance metrics, including sensitivity, specificity, area under the receiver operating curve ("AUC ROC"), and kappa statistics. P values less than 0.05 were considered to represent statistical significance. RESULTS: Anterior cruciate ligament tears were present in 45.7% (234/512) and absent in 54.3% (278/512) of the subjects. The DCNN had a sensitivity of 96.1%, which was not significantly different from the readers (97.5%-97.9%; all P ≥ 0.118), but significantly lower specificity of 93.1% (readers, 99.6%-100%; all P < 0.001) and "AUC ROC" of 0.935 (readers, 0.989-0.991; all P < 0.001) for the entire cohort. Subgroup analysis showed a significantly lower sensitivity, specificity, and "AUC ROC" of the DCNN for outside MRI (92.5%, 87.1%, and 0.898, respectively) than in-house MRI (99.0%, 94.4%, and 0.967, respectively) examinations (P = 0.026, P = 0.043, and P < 0.05, respectively). There were no significant differences in DCNN performance for 1.5-T and 3-T MRI examinations (all P ≥ 0.753, respectively). CONCLUSIONS: Deep Convolutional Neural Network performance of ACL tear diagnosis can approach performance levels similar to fellowship-trained full-time academic musculoskeletal radiologists at 1.5 T and 3 T; however, the performance may decrease with increasing MRI examination heterogeneity.

Deep convolutional neural network-based detection of meniscus tears: comparison with radiologists and surgery as standard of reference.

OBJECTIVE: To clinically validate a fully automated deep convolutional neural network (DCNN) for detection of surgically proven meniscus tears. MATERIALS AND METHODS: One hundred consecutive patients were retrospectively included, who underwent knee MRI and knee arthroscopy in our institution. All MRI were evaluated for medial and lateral meniscus tears by two musculoskeletal radiologists independently and by DCNN. Included patients were not part of the training set of the DCNN. Surgical reports served as the standard of reference. Statistics included sensitivity, specificity, accuracy, ROC curve analysis, and kappa statistics. RESULTS: Fifty-seven percent (57/100) of patients had a tear of the medial and 24% (24/100) of the lateral meniscus, including 12% (12/100) with a tear of both menisci. For medial meniscus tear detection, sensitivity, specificity, and accuracy were for reader 1: 93%, 91%, and 92%, for reader 2: 96%, 86%, and 92%, and for the DCNN: 84%, 88%, and 86%. For lateral meniscus tear detection, sensitivity, specificity, and accuracy were for reader 1: 71%, 95%, and 89%, for reader 2: 67%, 99%, and 91%, and for the DCNN: 58%, 92%, and 84%. Sensitivity for medial meniscus tears was significantly different between reader 2 and the DCNN (p = 0.039), and no significant differences existed for all other comparisons (all p ≥ 0.092). The AUC-ROC of the DCNN was 0.882, 0.781, and 0.961 for detection of medial, lateral, and overall meniscus tear. Inter-reader agreement was very good for the medial (kappa = 0.876) and good for the lateral meniscus (kappa = 0.741). CONCLUSION: DCNN-based meniscus tear detection can be performed in a fully automated manner with a similar specificity but a lower sensitivity in comparison with musculoskeletal radiologists.

Multi-variable models predicting specific patient-reported acute urinary symptoms after radiotherapy for prostate cancer: Results of a cohort study.

PURPOSE: A prospective trial started in 2010, aiming at developing models for urinary toxicity and erectile dysfunction after radiotherapy for prostate cancer. This analysis is finalised at highlighting correlations between clinical/dosimetric factors and acute urinary specific symptoms, as measured by single questions of the International Prostate Symptom Score (IPSS). MATERIALS/METHODS: IPSS was prospectively collected before and at the end of radiotherapy; absolute weekly bladder dose-surface histograms (DSHw) were chosen as dosimetric descriptors. Relevant clinical factors were prospectively gathered. Backward feature selection was used to identify variables to be included in logistic models for moderate-severe (scores⩾4) urinary symptoms. RESULTS: Complete data of 262 patients (120 conventional fractionation, 142 hypofractionation) were available. Smoking was a strong predictor for feeling of incomplete emptying, frequency, intermittency, urgency and straining; neoadjuvant hormonal therapy and use of antihypertensive drugs were risk factors for intermittency and weak stream, respectively. The baseline score was a major predictor for all symptoms with the exception of intermittency. DSHw were correlated to increased risk of frequency, intermittency, urgency and nocturia. Most models showed moderate-high discrimination (AUC≈0.60-0.79). CONCLUSIONS: Smoking and other clinical and dosimetric factors predict for specific moderate-severe acute urinary symptoms; baseline condition heavily modulated the risk in most endpoints.

Relationships between bladder dose-volume/surface histograms and acute urinary toxicity after radiotherapy for prostate cancer.

BACKGROUND AND PURPOSE: DUE01 is an observational study aimed at developing predictive models of genito-urinary toxicity of patients treated for prostate cancer with conventional (1.8-2Gy/fr, CONV) or moderate hypo-fractionation (2.35-2.7Gy/fr, HYPO). The current analysis focused on the relationship between bladder DVH/DSH and the risk of International Prostate Symptoms Score (IPSS)⩾15/20 at the end of radiotherapy. MATERIALS AND METHODS: Planning and relevant clinical parameters were prospectively collected, including DVH/DSH, LQ-corrected (DVHc/DSHc) and weekly (DVHw/DSHw) histograms. Best parameters were selected by the differences between patients with/without IPSS⩾15/20 at the end of radiotherapy. Logistic uni- and backward multi-variable (MVA) analyses were performed. RESULTS: Data of 247 patients were available (CONV: 116, HYPO: 131). Absolute DVHw/DSHw and DVHc/DSHc predicted the risk of IPSS⩾15 at the end of radiotherapy (n=77/247); an MVA model including baseline IPSS, anti-hypertensive, T stage, the absolute surface receiving ⩾8.5Gy/week and ⩾12.5Gy/week was developed (AUC=0.78, 95% CI: 0.72-0.83). Similar AUC values were found if replacing DSHw with DVHw/DVHc/DSHc parameters. The impact of dose-volume/surface parameters remained when excluding patients with baseline IPSS⩾15 and in HYPO. IPSS⩾20 at the end of radiotherapy (n=27/247) was mainly correlated to baseline IPSS and T stage. CONCLUSIONS: Although the baseline IPSS was the main predictor, constraining v8.5w<56cc and v12.5w<5cc may significantly reduce acute GU toxicity.