Search

Purpose: To create a system to enable the identification of histological variants of bladder cancer in a simple, efficient, and noninvasive manner. Material and methods: In this multicenter diagnostic study, we retrospectively collected basic information and CT images about the patients concerned from three hospitals. An interactive deep learning-based bladder cancer image segmentation framework was constructed using the Swin UNETR algorithm for further features extraction. Radiomic features and deep learning features were extracted for further stacking ensemble system construction. The segmentation model' performance was assessed by using Dice Similarity (Dice) metrics, Intersection Over Union (IOU), Sensitivity (SEN) and Specificity (SPE). To evaluate the system's performance, we used the Receiver Operating Characteristics (ROC) curve, the Accuracy Score (ACC) and Decision Curve Analysis (DCA). Results: 410 patients from one hospital were included in the training set, while 60 patients from two other hospitals were included in the test set. A total of 50 features comprising 46 radiomic features and 4 deep learning features were finally retained for further stacking ensemble model building. The interactive segmentation model and system exhibited excellent performance in both training (Dice = 0.78, IOU = 0.65, SEN = 0.83, SPE = 1.00, AUC = 0.940, ACC = 0.868) and testing datasets (Dice = 0.80, IOU = 0.67, SEN = 0.89, SPE = 1.00, AUC = 0.905, ACC = 0.900). Conclusion: We successfully constructed a stacking ensemble machine learning model for early, non-invasive identification of histological variants in bladder cancer which will help urologists make clinical decisions. [ABSTRACT FROM AUTHOR]

Objectives: To propose a reliable and standard 3D assessment method to analyze the effect of weight-bearing (WB) status on the location of patella and clarify the diagnostic performance of 3D parameters for recurrent patellar dislocation (RPD) in WB and non-weight-bearing (NWB) conditions. Methods: Sixty-five knees of RPD patients and 99 knees of controls were included. Eight landmarks, two lines and a coordinate system were defined on 3D bone models of knees based on weight-bearing CT and non-weight-bearing CT. The shift and tilt of patella in three orthogonal axes (Xshift, Yshift, Zshift, Xtilt, Ytilt, Ztilt) were evaluated. Results: Xshift and Yshift were significantly higher, Zshift, Xtilt and Ytilt were significantly lower in WB condition than NWB condition (p < 0.001, p < 0.001, p = 0.001, p = 0.002, p = 0.010). In both WB and NWB conditions, Xshift, Yshift and Ztilt were significantly higher, and Xtilt was significantly lower in the RPD group than the control group (WB/NWB: p < 0.001/p = 0.002, p < 0.001/p = 0.001, p < 0.001/p < 0.001, p < 0.001/p = 0.009). In WB condition, Zshift and Ytilt were significantly higher in the RPD group than the control group (p = 0.011, p < 0.001). Ztilt had the best diagnostic performance for RPD in both WB and NWB conditions, with AUC of 0.887 (95% CI: 0.828, 0.946) and 0.885 (95% CI: 0.822, 0.947), respectively. Conclusions: The 3D measurement method reliably and comprehensively reflected the relative spatial position relationship of the patellofemoral joint. It can be applied to the 3D preoperative planning of patellofemoral procedures. In addition, patellofemoral evaluation under the WB condition was essential to detect subtle underlying risk factors for RPD, with axial lateral patellar tilt being the best predictor. Critical relevance statement: This 3D measurement method under weight-bearing conditions contributes to comprehensively describing the relative spatial position of the patellofemoral joint in a standardized way and can be applied to preoperative evaluation for recurrent patellar dislocation. Key Points: Patellofemoral alignment is a 3D problem, and the accuracy of 2D parameters has been questioned. 3D measurement was reliable and comprehensively reflected relative spatial relationships of the patellofemoral joint. 3D measurements under weight-bearing condition help preoperative evaluation for RPD. [ABSTRACT FROM AUTHOR]

Purpose: The aim of this study was to develop and validate a prediction model for classification of pulmonary nodules based on preoperative CT imaging. Materials and methods: A data set of Centers 1 (training set: 2633; internal testing set: 1129); Center 2 and Center 3 (external testing set: 218) of patients with pulmonary nodule cases was retrospectively collected. Handcrafted features were extracted from noncontrast chest CT scans by three senior radiologists. A total of 22 clinically handcrafted parameters (age, gender, L-RADS, and PNI-GARS et al.) were used to construct machine learning models (random forest, gradient boosting, and explainable boosting) for the classification of preoperative pulmonary nodules, and the parameters of the model were adjusted to achieve optimal performance. To evaluate the prediction capacity of each model. Both 5-fold cross-validation and 10-fold cross-validation were used to test the robustness of the models. Results: The explainable boosting model had the best performance on our constructed data. The model achieves an accuracy of 89.9%, a precision of 97.48%, a specificity of 89.5%, a sensitivity of 91.1%, and an AUC of 90.3%. In human-machine comparison, the AUC of machine learning models (90.4%, 95% CI: 85.5%–94.8%) was significantly improved compared to radiologists (60%, 95% CI: 50%–71.4%). Conclusions: The explainable boosting model exhibited superior performance on our dataset, achieving high accuracy and precision in the diagnosis of pulmonary nodules compared to experienced radiologists. [ABSTRACT FROM AUTHOR]

The clinical magnetic resonance scanner (field strength ≤ 3.0 T) has limited efficacy in the high-resolution imaging of experimental mice. This study introduces a novel magnetic resonance micro-coil designed to enhance the signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR), thereby improving high-resolution imaging in experimental mice using clinical magnetic resonance scanners. Initially, a phantom was utilized to determine the maximum spatial resolution achievable by the novel micro-coil. Subsequently, 12 C57BL/6JGpt mice were included in this study, and the novel micro-coil was employed for their scanning. A clinical flexible coil was selected for comparative analysis. The scanning methodologies for both coils were consistent. The imaging clarity, noise, and artifacts produced by the two coils on mouse tissues and organs were subjectively evaluated, while the SNR and CNR of the brain, spinal cord, and liver were objectively measured. Differences in the images produced by the two coils were compared. The results indicated that the maximum spatial resolution of the novel micro-coil was 0.2 mm. Furthermore, the subjective evaluation of the images obtained using the novel micro-coil was superior to that of the flexible coil (p < 0.05). The SNR and CNR measurements for the brain, spinal cord, and liver using the novel micro-coil were significantly higher than those obtained with the flexible coil (p < 0.001). Our study suggests that the novel micro-coil is highly effective in enhancing the image quality of clinical magnetic resonance scanners in experimental mice. [ABSTRACT FROM AUTHOR]