Search

Abstract Understanding the mechanical characteristics of rocks when subjected to direct tension is crucial for assessing the stability of rock formations. Within the scope of this research, a series of tests was conducted using Tage tuff to assess the deformation behavior and crack extension of rock under direct tension. The axial, lateral, and shear strain fields as well as crack propagation, localized deformation behavior, and failure mode of the rocks were analyzed using three-dimensional digital image correlation method. The results showed that the axial strain fields on the specimen surface were heterogeneous, with different locations and localization occurring in the pre-peak stage, which was similar to the evolution of shear strain, whereas the lateral strain only showed slight changes. The crack extension direction was inferred, indicating that both tensile and shear stress occurred in the tests. Furthermore, different stress–strain responses were observed for the inside and outside of the localized bands. Then, the surface patterns of specimen failure were scanned and analyzed to assess the failure mode and residual strength of the specimen under direct tensile stress. Finally, the results of direct tension, uniaxial compression, and Brazilian split tests for Tage tuff were compared, and the complete stress–strain curve of uniaxial tension (UT) was simulated using a nonlinear-variable-compliance constitutive equation. This study provides a deeper understanding into the damage behavior of rocks under direct tension.

Abstract The frozen section (FS) diagnoses of pathology experts are used in China to determine whether sentinel lymph nodes of breast cancer have metastasis during operation. Direct implementation of a deep neural network (DNN) in clinical practice may be hindered by misdiagnosis of the algorithm, which affects a patient's treatment decision. In this study, we first obtained the prediction result of the commonly used patch-DNN, then we present a relative risk classification and regression tree (RRCART) to identify the misdiagnosed whole-slide images (WSIs) and recommend them to be reviewed by pathologists. Applying this framework to 2362 WSIs of breast cancer lymph node metastasis, test on frozen section results in the mean area under the curve (AUC) reached 0.9851. However, the mean misdiagnosis rate (0.0248), was significantly higher than the pathologists’ misdiagnosis rate (p 0.01). However, the other low-accuracy group included most of the misdiagnoses of DNN models. Our research shows that the misdiagnosis from deep learning model can be further enriched by our method, and that the low-accuracy WSIs must be selected for pathologists to review and the high-accuracy ones may be ready for pathologists to give diagnostic reports.