The success of image classification depends on copious annotated images for training. Annotating histopathology images is costly and laborious. Although several successful self-supervised representation learning approaches have been introduced, they are still insufficient to consider the unique characteristics of histopathology images. In this work, we propose the novel histopathology-oriented self-supervised representation learning framework (HistoSSL) to efficiently extract representations from unlabeled histopathology images at three levels: global, cell, and stain. The model transfers remarkably to downstream tasks: colorectal tissue phenotyping on the NCTCRC dataset and breast cancer metastasis recognition on the CAMELYON16 dataset. HistoSSL achieved higher accuracies than state-of-the-art self-supervised learning approaches, which proved the robustness of the learned representations.
The emergence of cloud computing and big data has rapidly increased the deployment of small and mid-sized data centers. Enterprises and cloud providers require an agile network among these data centers to empower application reliability and flexible scalability. We present a software-defined inter data center network to enable on-demand scale out of data centers on a metro-scale optical network. The architecture consists of a combined space/wavelength switching platform and a Software-Defined Networking (SDN) control plane equipped with a wavelength and routing assignment module. It enables establishing transparent and bandwidth-selective connections from L2/L3 switches, on-demand. The architecture is evaluated in a testbed consisting of 3 data centers, 5–25 km apart. We successfully demonstrated end-to-end bulk data transfer and Virtual Machine (VM) migrations across data centers with less than 100 ms connection setup time and close to full link capacity utilization.
Although surgical resection is the best option for treating gliomas, it might be difficult to minimize the harm done to healthy brain regions. As a result, segmenting brain tumors in medical image analysis has become essential. A number of segmentation investigations utilizing CNNs have lately demonstrated promising performance thanks to the advancement of image equipment and deep learning techniques. In this study, we propose the SCAR U-Net, an end-to-end 3D residual U-Net model incorporating attention mechanisms for the segmentation of brain tumors. The SCAR U-Net employs channel and spatial attention processes and has a 3D U-Net architecture with residual blocks. We evaluate the model on a subset of the BraTS 2021 dataset. And the model outperforms the baseline significantly by ET, TC, and WT in the test set. Finally, we use ablation tests to confirm the beneficial effects of the residual connections and the attention modules.