Search

In recent years, deep convolutional neural networks (DCNNs) have shown promising performance in medical image analysis, including breast lesion classification in 2D ultrasound (US) images. Despite the outstanding performance of DCNN solutions, explaining their decisions remains an open investigation. Yet, the explainability of DCNN models has become essential for healthcare systems to accept and trust the models. This paper presents a novel framework for explaining DCNN classification decisions of lesions in ultrasound images using the saliency maps linking the DCNN decisions to known cancer characteristics in the medical domain. The proposed framework consists of three main phases. First, DCNN models for classification in ultrasound images are built. Next, selected methods for visualization are applied to obtain saliency maps on the input images of the DCNN models. In the final phase, the visualization outputs and domain-known cancer characteristics are mapped. The paper then demonstrates the use of the framework for breast lesion classification from ultrasound images. We first follow the transfer learning approach and build two DCNN models. We then analyze the visualization outputs of the trained DCNN models using the EGrad-CAM and Ablation-CAM methods. We map the DCNN model decisions of benign and malignant lesions through the visualization outputs to the characteristics such as echogenicity, calcification, shape, and margin. A retrospective dataset of 1298 US images collected from different hospitals is used to evaluate the effectiveness of the framework. The test results show that these characteristics contribute differently to the benign and malignant lesions' decisions. Our study provides the foundation for other researchers to explain the DCNN classification decisions of other cancer types. [ABSTRACT FROM AUTHOR]

This paper explores the adaptations Buddhism has made to digital payment methods in the context of the Chinese mainland. To provide the audience with a relatively comprehensive understanding of the general context in which the new method of donation is applied, this paper first introduces the development and digital landscape of the internet and third-party payments in the Chinese mainland. Then, statistical analysis is used to make large-scale claims by analyzing 1328 donation records made to Donglin Monastery in Mount Lu with the statistical software SPSS to determine whether the digital donation method is linked to the purpose of donations, or the amount of money being donated, and to what extent it substitutes for traditional donation methods. [ABSTRACT FROM AUTHOR]

Supramolecular polymer networks (SPNs) demonstrate great potential in the development of smart materials owing to their attractive dynamic properties. However, as they suffer from the inherent weak bonding of most noncovalent cross‐links, it remains a significant challenge to construct SPNs with outstanding mechanical performance. Herein, we exploit the cryptand/paraquat host‐guest recognition motifs as cross‐links to prepare a class of highly strong and tough SPNs. Unlike those supramolecular cross‐links with relatively weak binding abilities, the cryptand‐based host‐guest interactions have a high association constant and steady complexing structure, which effectively stabilizes the network and resists mechanical deformation under external force. Such favorable structural stability endows our SPNs with greatly enhanced mechanical performance, compared with the control‐1 cross‐linked by the weakly complexed crown ether/secondary ammonium salt motif (tensile strength: 21.1±0.5 vs 2.8±0.1 MPa; Young's modulus: 102.6±4.8 vs 2.1±0.3 MPa; toughness: 90.4±2.0 vs 10.8±0.6 MJ m−3). Moreover, our SPNs also retain abundant dynamic properties including good abilities in energy dissipation, reprocessability, and stimuli‐responsiveness. These findings provide novel insights into the preparation of SPNs with enhanced mechanical properties, and promote the development of high‐performance intelligent supramolecular materials. [ABSTRACT FROM AUTHOR]

I Future Internet i was able to uphold its high standards for published papers due to the outstanding efforts of our reviewers. High-quality academic publishing is built on rigorous peer review. [Extracted from the article]

Modulation and Modeling of Three-Dimensional Nanowire Assemblies Targeting Gas Sensors with High Response and Reliability (Adv. Funct. Gas sensors, metal oxides, nanoarchitecture, nano-transfer printing Keywords: gas sensors; metal oxides; nanoarchitecture; nano-transfer printing EN gas sensors metal oxides nanoarchitecture nano-transfer printing 1 1 1 03/04/22 20220308 NES 220308 B Gas Sensors b In article number 2108891, Harry L. Tuller, Yeon Sik Jung, and co-workers systematically engineer 3D metal oxide nanowire assemblies to achieve superior gas response, conductance, and signal stability compared to random nanowire aggregates. [Extracted from the article]