Search

X-ray excited persistent luminescence (XEPL) imaging has attracted increasing attention in biomedical imaging due to elimination of autofluorescence, high signal-to-noise ratio and repeatable activation with high penetration. However, optical imaging still suffers from limited for high spatial resolution. Methods: Herein, we report Mn3+-rich manganese oxide (MnOx)-coated chromium-doped zinc gallogermanate (ZGGO) nanoparticles (Mn-ZGGOs). Enhanced XEPL and magnetic resonance (MR) imaging were investigated by the decomposition of MnOx shell in the environment of tumors. We also evaluated the tumor cell-killing mechanism by detection of reactive oxygen (ROS), lipid peroxidation and mitochondrial membrane potential changes in vitro. Furthermore, the in vivo biodistribution, imaging and therapy were studied by U87MG tumor-bearing mice. Results: In the tumor region, the MnOx shell is quickly decomposed to produce Mn3+ and oxygen (O2) to directly generate singlet oxygen (1O2). The resulting Mn2+ transforms endogenous H2O2 into highly toxic hydroxyl radical (·OH) via a Fenton-like reaction. The Mn2+ ions and ZGGOs also exhibit excellent T1-weighted magnetic resonance (MR) imaging and ultrasensitive XEPL imaging in tumors. Conclusion: Both the responsive dual-mode imaging and simultaneous self-supplied O2 for the production of 1O2 and oxygen-independent ·OH in tumors allow for more accurate diagnosis of deep tumors and more efficient inhibition of tumor growth without external activation energy.

X-ray excited persistent luminescence (XEPL) imaging has attracted increasing attention in biomedical imaging due to elimination of autofluorescence, high signal-to-noise ratio and repeatable activation with high penetration. However, optical imaging still suffers from limited for high spatial resolution. Methods: Herein, we report Mn3+-rich manganese oxide (MnOx)-coated chromium-doped zinc gallogermanate (ZGGO) nanoparticles (Mn-ZGGOs). Enhanced XEPL and magnetic resonance (MR) imaging were investigated by the decomposition of MnOx shell in the environment of tumors. We also evaluated the tumor cell-killing mechanism by detection of reactive oxygen (ROS), lipid peroxidation and mitochondrial membrane potential changes in vitro. Furthermore, the in vivo biodistribution, imaging and therapy were studied by U87MG tumor-bearing mice. Results: In the tumor region, the MnOx shell is quickly decomposed to produce Mn3+ and oxygen (O2) to directly generate singlet oxygen (1O2). The resulting Mn2+ transforms endogenous H2O2 into highly toxic hydroxyl radical (·OH) via a Fenton-like reaction. The Mn2+ ions and ZGGOs also exhibit excellent T1-weighted magnetic resonance (MR) imaging and ultrasensitive XEPL imaging in tumors. Conclusion: Both the responsive dual-mode imaging and simultaneous self-supplied O2 for the production of 1O2 and oxygen-independent ·OH in tumors allow for more accurate diagnosis of deep tumors and more efficient inhibition of tumor growth without external activation energy.

Deep neural networks (DNNs) have demonstrated their great potential in recent years, exceeding the per-formance of human experts in a wide range of applications. Due to their large sizes, however, compressiontechniques such as weight quantization and pruning are usually applied before they can be accommodated onthe edge. It is generally believed that quantization leads to performance degradation, and plenty of existingworks have explored quantization strategies aiming at minimum accuracy loss. In this paper, we argue thatquantization, which essentially imposes regularization on weight representations, can sometimes help toimprove accuracy. We conduct comprehensive experiments on three widely used applications: fully con-nected network (FCN) for biomedical image segmentation, convolutional neural network (CNN) for imageclassification on ImageNet, and recurrent neural network (RNN) for automatic speech recognition, and experi-mental results show that quantization can improve the accuracy by 1%, 1.95%, 4.23% on the three applicationsrespectively with 3.5x-6.4x memory reduction., 11 pages, 3 figures, 10 tables, accepted by the ACM Journal on Emerging Technologies in Computing Systems (JETC)