No abstract

No abstract

Bacterial cellulose (BC) is a natural biomaterial with unique properties suitable for tissue engineering applications, but it has not yet been used for preparing nerve conduits to repair peripheral nerve injuries. The objectives of this study were to prepare and characterize the Kampuchea-synthesized bacterial cellulose (KBC) and further evaluate the biocompatibility of KBC with peripheral nerve cells and tissues in vitro and in vivo. KBC membranes were composed of interwoven ribbons of about 20-100 nm in width, and had a high purity and the same crystallinity as that of cellulose Iα. The results from light and scanning electron microscopy, MTT assay, flow cytometry, and RT-PCR indicated that no significant differences in the morphology and cell function were observed between Schwann cells (SCs) cultured on KBC membranes and glass slips. We also fabricated a nerve conduit using KBC, which was implanted into the spatium intermusculare of rats. At 1, 3, and 6 weeks post-implantation, clinical chemistry and histochemistry showed that there were no significant differences in blood counts, serum biochemical parameters, and tissue reactions between implanted rats and sham-operated rats. Collectively, our data indicated that KBC possessed good biocompatibility with primary cultured SCs and KBC did not exert hematological and histological toxic effects on nerve tissues in vivo.

No abstract

Trained with the standard cross entropy loss, deep neural networks can achieve great performance on correctly labeled data. However, if the training data is corrupted with label noise, deep models tend to overfit the noisy labels, thereby achieving poor generation performance. To remedy this issue, several loss functions have been proposed and demonstrated to be robust to label noise. Although most of the robust loss functions stem from Categorical Cross Entropy (CCE) loss, they fail to embody the intrinsic relationships between CCE and other loss functions. In this paper, we propose a general framework dubbed Taylor cross entropy loss to train deep models in the presence of label noise. Specifically, our framework enables to weight the extent of fitting the training labels by controlling the order of Taylor Series for CCE, hence it can be robust to label noise. In addition, our framework clearly reveals the intrinsic relationships between CCE and other loss functions, such as Mean Absolute Error (MAE) and Mean Squared Error (MSE). Moreover, we present a detailed theoretical analysis to certify the robustness of this framework. Extensive experimental results on benchmark datasets demonstrate that our proposed approach significantly outperforms the state-of-the-art counterparts.

Poor osseointegration and infection resulting from implants are serious medical issues, and it is not straightforward to manufacture implants that can simultaneously address both of these problems. In this study, we produced coatings containing titania nanotubes (TiO2 -NTs) incorporated with zinc (NT-Zn) on Ti substrates by anodization and hydrothermal treatment. The zinc content was controlled by varying the duration of the hydrothermal treatment. The NT-Zn implants not only exhibited improved bone formation (shown by both in vitro and in vivo studies), which enhances osseointegration between bone and implant, but also inhibited growth of bacteria. The cytotoxicity of locally high concentrations of zinc in the NT-Zn3h specimens observed during in vitro studies was mitigated by the effects of dilution in vivo.

No abstract

No abstract