Background: Continuous contrast-enhanced ultrasound (CEUS) video is a challenging direction for radiomics research. We aimed to evaluate machine learning (ML) approaches with radiomics combined with the XGBoost model and a convolutional neural network (CNN) for discriminating between benign and malignant lesions in CEUS videos with a duration of more than 1 min.Methods: We gathered breast CEUS videos of 109 benign and 81 malignant tumors from two centers. Radiomics combined with the XGBoost model and a CNN was used to classify the breast lesions on the CEUS videos. The lesions were manually segmented by one radiologist. Radiomics combined with the XGBoost model was conducted with a variety of data sampling methods. The CNN used pretrained 3D residual network (ResNet) models with 18, 34, 50, and 101 layers. The machine interpretations were compared with prospective interpretations by two radiologists. Breast biopsies or pathological examinations were used as the reference standard. Areas under the receiver operating curves (AUCs) were used to compare the diagnostic performance of the models.Results: The CNN model achieved the best AUC of 0.84 on the test cohort with the 3D-ResNet-50 model. The radiomics model obtained AUCs between 0.65 and 0.75. Radiologists 1 and 2 had AUCs of 0.75 and 0.70, respectively.Conclusions: The 3D-ResNet-50 model was superior to the radiomics combined with the XGBoost model in classifying enhanced lesions as benign or malignant on CEUS videos. The CNN model was superior to the radiologists, and the radiomics model performance was close to the performance of the radiologists.

Objective. We aimed to evaluate and compare the association between globulin to albumin ratio (GAR) and globulin to prealbumin ratio (GPR) and 3-month functional prognosis of acute ischemic stroke (AIS) patients receiving intravenous thrombolysis therapy. Methods.234 AIS patients undergoing intravenous thrombolysis were retrospectively enrolled with acute ischemic stroke from February 2016 to October 2019. Blood sample was collected within 24 h after admission. Poor outcome was defined as the modified and a favorable outcome as . Severe stroke was defined as the on admission. Student’s -test, Mann–Whitney test, Chi-square test, logistics’ regression analysis, and receiver operating characteristic (ROC) analysis were performed. Results. Patients with poor functional outcome had higher GAR and GPR levels compared with favorable functional group ( , , respectively). Severe stroke was also associated with these two increasing variables. After adjustment for confounding factors, multivariate logistic regression analysis indicated that GPR was an independent indicator predictor of AIS. Conclusions. The 24 h GPR level can predict the 3-month functional outcome in AIS patients accepting recombinant tissue plasminogen activator (r-tPA) intravenous thrombosis.