Citations

ABSTRACTBackgroundDeep learning (DL) models for accurate renal tumor characterization may benefit from multi‐center datasets for improved generalizability; however, data‐sharing constraints necessitate privacy‐preserving solutions like federated learning (FL).PurposeTo assess the performance and reliability of FL for renal tumor segmentation and classification in multi‐institutional MRI datasets.Study TypeRetrospective multi‐center study.PopulationA total of 987 patients (403 female) from six hospitals were included for analysis. 73% (723/987) had malignant renal tumors, primarily clear cell carcinoma (n = 509). Patients were split into training (n = 785), validation (n = 104), and test (n = 99) sets, stratified across three simulated institutions.Field Strength/SequenceMRI was performed at 1.5 T and 3 T using T2‐weighted imaging (T2WI) and contrast‐enhanced T1‐weighted imaging (CE‐T1WI) sequences.AssessmentFL and non‐FL approaches used nnU‐Net for tumor segmentation and ResNet for its classification. FL‐trained models across three simulated institutional clients with central weight aggregation, while the non‐FL approach used centralized training on the full dataset.Statistical TestsSegmentation was evaluated using Dice coefficients, and classification between malignant and benign lesions was assessed using accuracy, sensitivity, specificity, and area under the curves (AUCs). FL and non‐FL performance was compared using the Wilcoxon test for segmentation Dice and Delong's test for AUC (p < 0.05).ResultsNo significant difference was observed between FL and non‐FL models in segmentation (Dice: 0.43 vs. 0.45, p = 0.202) or classification (AUC: 0.69 vs. 0.64, p = 0.959) on the test set. For classification, no significant difference was observed between the models in accuracy (p = 0.912), sensitivity (p = 0.862), or specificity (p = 0.847) on the test set.Data ConclusionFL demonstrated comparable performance to non‐FL approaches in renal tumor segmentation and classification, supporting its potential as a privacy‐preserving alternative for multi‐institutional DL models.Evidence Level4.Technical EfficacyStage 2.

Abstract Background: Limited information existed about the distributions of microbial etiology and drug-resistance with severe community-acquired pneumonia (CAP) in China, partly attributing to the antibiotics abuse in rearing livestock for Chinese farmers in recent decades. Moreover, the relevant guidelines have not been updated for at least a decade.Methods: A prospective observational study was carried out in Zhongshan Hospital (Shanghai, China). All consecutive cases of severe CAP diagnosed in Respiratory intensive care unit (ICU) and Emergency ICU were included. We assessed the microbial etiology features and multidrug-resistance (MDR) status. The risk factors were used by multivariate analyses for outcomes.Results: One hundred and twelve adults with severe CAP in ICUs, mean age of 65.7±15.1 years were studied. Etiological diagnoses were established in 76 (67.9%) cases; Klebsiella pneumoniae was the most frequent pathogen in 27 (35.5%) cases followed by Acinetobacter baumannii (n=15, 19.7%) and Methicillin-resistant Staphylococcus aureus (MRSA) (n=9, 11.8%); While Streptococcus pneumoniae (n=3, 3.9%) was ranked the 6 th in the study. The majority of pathogens were the MDR and the 30-days mortality rate was 43.8% in severe CAP patients. Klebsiella pneumoniae (OR, 6.88; 95%CI, 1.13-41.83), Sequential organ failure assessment (SOFA)≥5(OR, 12.30; 95%CI, 1.79-84.55) was associated to death.Conclusions: The study indicates a change in the usual pattern of microbiol etiology of severe CAP and the MDR emerged in the majority of pathogens over a decade in eastern China. Enhanced control of using antibiotics to reduce the acquired resistance is urgent.