Since zinc oxide (ZnO) nanoparticles (NPs) have been widely applied, the nano community and the general public have paid great attention to the toxicity of ZnO NPs. We detected 20-nm ZnO NPs biotoxicity following nasal exposure utilizing the non-invasive and real-time magnetic resonance imaging (MRI) technique. MR images were scanned in the rat olfactory epithelium (OE) and olfactory bulb (OB) on a 4.7 T scanner following the treatment (as early as 1 day and up to 21 days after), and the histological changes were evaluated. The influence of the size of the ZnO NPs and chemical components was also investigated. Our study revealed that 20-nm ZnO NPs induced obvious structural disruption and inflammation in the OE and OB at the acute stage. The results suggest that the real-time and non-invasive advantages of MRI allow it to observe and assess, directly and dynamically, the potential toxicity of long-term exposure to ZnO NPs in the olfactory system. These findings indicate the size-dependent toxicity of ZnO NPs with respect to the olfactory bulb. Further study is needed to reveal the mechanism behind ZnO NPs’ toxicity.

In the task of liver image segmentation, due to the complex edges and internal structure of liver images, as well as the presence of adjacent organs, conventional medical image segmentation networks have exhibited local deficiencies in handling long-range dependencies and multi-scale information. These deficiencies may impact the accuracy of liver segmentation models. To address the small receptive fields of convolutional operations in previous medical image segmentation models and the information loss issue in the skip connections of the TransUNet model, this paper introduces Axial Attention and residual connections into the skip connections of the TransUNet model. Axial Attention weights the importance of different dimensions, better capturing and enhancing liver edges and other targets. Additionally, the use of residual connections effectively alleviates gradient vanishing and information loss issues. Experimental results demonstrate that the improved TransUNet model proposed in this paper achieves dice and IoU evaluation metrics of 97.2% and 94.0%, respectively, on the LiTS17 dataset, representing an improvement of 1.6 percentage points and 2.1 percentage points in Dice and IoU metrics, respectively, compared to the original TransUNet model. Its liver segmentation performance surpasses that of the original TransUNet model, and the improved TransUNet model can more accurately identify liver edges, enhancing feature representation and information flow, and providing more refined segmentation results.