Autologous transplantation remains the golden standard for peripheral nerve repair.However, many drawbacks, such as the risk of reoperation or nerve injury remain associated with this method. To date, commercially available artificial nerve conduits comprise hollow tubes. By providing physical guiding and biological cues, tissue engineered conduits are promising for bridging peripheral nerve defects. The present study focuses on the preparation of artificial composite nerve conduits by 3D bio-printing. 3D-printed molds with a tubular cavity were filled with an Engelbreth-Holm-Swarm (EHS) Hydrogel mimicking the extracellular matrix (ECM) basement membrane. Chemically cross-linked gelatin methacryloyl (GelMA) was used to form the conduit backbone, while EHS Hydrogels improved nerve fiber growth while shortening repair time. Statistical significant difference had been found between the blank conduit and the composite conduit group on compound muscle action potential after 4 months. On the other hand, results between the composite conduit group and the autograft group were of no statistical differences.All results above showed that the composite conduit filled with EHS Hydrogel can promote the repair of peripheral nerve and may become a promising way to treat peripheral nerve defects.3D bioprint, chemical cross-link, ECM, GelMA, nerve repair Conduits were produced as described earlier (Hu et al., 2016). Briefly, the mold (Figure 1a,b) was modeled with the use of SolidWorks to create a tubular cavity prior to 3D printing. Based on the size of rat sciatic nerves, the inner and outer diameters were 2 and 4 mm, respectively (Zhao et al., 2016). Five percentage of GelMA (SE-3DP-0200, StemEasy Biotech, China) aqueous solution was prepared at 60 C. After complete dissolution, the solution was cooled to 15 C prior to the addition of 0.5% APS and 0.1% TEMED. The mixture was then poured into the mold and rapidly placed at −20 C. After 24 hr of solidification, the conduit was defrosted and unmolded, followed by intensive rinsing and freeze-drying.For EHS Hydrogel (SE-EHS-0100, StemEasy Biotech, China) filling, conduits were placed on precooled petri-dishes and 0.1 ml of EHS Hydrogel was poured into the cavity. Conduits were then immediately placed in an incubator. After 5 min, the EHS Hydrogel displayed a jellylike texture and the composite conduit was considered ready to use.According to the conventional method, the blank GelMA conduit was examined by scanning electron microscopy (SEM, Phenom ProX, Thermo fisher), and the images were analyzed by ImageJ to calculate the porosity. The blank conduits were immersed in saline in 37 C, and the swelling ratio was calculated by weighing before and after at different time points and calculated by the following formula (Zhao et al., 2016):Swelling ratio = wet weight− dry weight dry weight × 100%The mechanical properties of GelMA blank conduits were also evaluated by both compression and tensile tests using universal testing systems (5969, Instron, United States) equipped with a 50 N loa...

The recent Natural Language Processing techniques have been refreshing the stateof-the-art performance at an incredible speed. Training huge language models is therefore an imperative demand in both industry and academy. However, the huge models impose challenges to both hardware and software. Graphical processing units (GPUs) are iterated frequently to meet the exploding demand, and a variety of ASICs like TPUs are spawned. However, there is still a tension between the fast growth of the extremely huge models and fact that Moore's law is approaching the end. To this end, many model parallelism techniques are proposed to distribute the model parameters to multiple devices, so as to alleviate the tension on both memory and computation. Our work is the first to introduce a 3-dimensional model parallelism for expediting huge language models. By reaching a perfect load balance, our approach presents smaller memory and communication cost than existing state-of-the-art 1-D and 2-D model parallelism. Our experiments on 64 TACC's V100 GPUs show that our 3-D parallelism outperforms the 1-D and 2-D parallelism with 2.32X and 1.57X speedup, respectively.

Photodynamic therapy (PDT) is a non-invasive method for cancer treatment that relies on the generation of excess reactive oxygen species (ROS), upon excitation of photosensitizer (PS), to eradicate tumor cells. However, the overexpress of endogenous antioxidants in tumor cells will eliminate the ROS and restrict the therapeutic efficacy of PDT. Herein, a novel type of PS was developed by conjugating cinnamaldehyde (CA), a kind of oxidative stress amplified agent, with porpholactam through a hydrazone bond. The new PS retains the photophysical properties of porpholactam, which displays high singlet oxygen quantum yield for the PDT function. The results of in vitro experiments performed including ROS assay and the cytotoxicity in cancer cells suggest that the rational design of the novel porpholactam-CA derivatives result in enhanced ROS generation upon irradiation, providing a possible approach to achieve enhanced therapeutic effects in PDT.

No abstract