Using Vision Transformer for high robustness and generalization in predicting EGFR mutation status in lung adenocarcinoma.

BACKGROUND: Lung adenocarcinoma is a common cause of cancer-related deaths worldwide, and accurate EGFR genotyping is crucial for optimal treatment outcomes. Conventional methods for identifying the EGFR genotype have several limitations. Therefore, we proposed a deep learning model using non-invasive CT images to predict EGFR mutation status with robustness and generalizability. METHODS: A total of 525 patients were enrolled at the local hospital to serve as the internal data set for model training and validation. In addition, a cohort of 30 patients from the publicly available Cancer Imaging Archive Data Set was selected for external testing. All patients underwent plain chest CT, and their EGFR mutation status labels were categorized as either mutant or wild type. The CT images were analyzed using a self-attention-based ViT-B/16 model to predict the EGFR mutation status, and the model's performance was evaluated. To produce an attention map indicating the suspicious locations of EGFR mutations, Grad-CAM was utilized. RESULTS: The ViT deep learning model achieved impressive results, with an accuracy of 0.848, an AUC of 0.868, a sensitivity of 0.924, and a specificity of 0.718 on the validation cohort. Furthermore, in the external test cohort, the model achieved comparable performances, with an accuracy of 0.833, an AUC of 0.885, a sensitivity of 0.900, and a specificity of 0.800. CONCLUSIONS: The ViT model demonstrates a high level of accuracy in predicting the EGFR mutation status of lung adenocarcinoma patients. Moreover, with the aid of attention maps, the model can assist clinicians in making informed clinical decisions.

Assessment of pulmonary toxicity of potential antioxidant drug PEGylated nanoceria after intratracheal instillation in rats.

Cerium oxide (CeO2 ) nanoparticles have unique redox properties and exert excellent antioxidant effects in the biological environment. In recent years, many researchers have focused on the CeO2 nanoparticles as an effective antioxidant drug in the prevention and treatment of various diseases. However, the toxicity of CeO2 nanoparticles in vivo remains controversial and still needs intensive research. Therefore, the objective of this study is to investigate the pulmonary and systemic toxicity in rats after 14 days of exposure to the PEGylated CeO2 nanoparticles (abbreviated as CNPs; exposure dose of 2, 10, or 20 mg/kg) through a single intratracheal instillation (IT). We assessed the indicators of lung injury and the pathological damage degree of lung tissue. The bronchoalveolar lavage fluid (BALF) analysis and lung histopathology revealed the occurrence of slight pulmonary inflammation in the 20-mg/kg experimental group rats. However, the inflammation factors in the lung tissue of every group rats did not significantly increase, and the levels of superoxide dismutase (SOD) and glutathione (GSH) in lung tissue homogenate rose considerably in the experimental groups. Collectively, these results indicated that pulmonary exposure by the high dose of CNPs could induce mild pulmonary inflammation but did not cause severe systemic toxicity. Moreover, we speculate that the mechanism of pulmonary toxicity of CNPs in rats was due to the autophagic death of healthy lung epithelial cells mediated by endoplasmic reticulum stress. Our results implicate that CNPs can be safely used as an antioxidant drug for the oxidative stress pulmonary diseases.

Acute hepatotoxicity of multimodal targeted imaging contrast agent NaLuF4:Gd,Yb,Er-PEG/PEI-FA in mice.

In the past few decades, upconversion nanoparticles (abbreviated as UCNPs) have been more widely applied in the biomedical fields, such as in vitro and in vivo upconversion fluorescent bioimaging, photodynamic therapy, biological macromolecular detection, imaging mediated drug delivery and so on. But meanwhile, there is still not much research on the acute toxicity of upconversion nanoparticles in vivo, such as acute hepatotoxicity. In this work, we studied the in vivo biodistribution and acute hepatotoxicity of multimodal targeted contrast agent NaLuF4:Gd,Yb,Er-PEG/PEI-FA nanoprobe, which were synthesized by the solvothermal method and modified with Polyethylene glycol (PEG), Polyetherimide (PEI), folic acid (FA) on the surface. The acute hepatotoxicity in mice was systematically assessed after tail vein injection of different concentration of UCNPs. The results showed that NaLuF4:Gd,Yb,Er-PEG/PEI-FA nanoparticles with an average diameter of 44.5 ± 10.4 nm, and three typical upconversion fluorescence emission bands at 520 nm, 540 nm and 660 nm under the excitation of 980 nm laser. In vivo distribution experiments results demonstrated that approximately 87% of UCNPs injected through the tail vein accumulate in the liver. In the acute hepatotoxicity test, the intravenously injection dose of UCNPs was 10, 40, 70 and 100 mg/kg, respectively. The body weight, blood routine, serum biochemistry, histomorphology and liver oxidative stress were detected and observed no significant acute hepatotoxicity damage under the injection dose of 100 mg/kg. In conclusion, NaLuF4:Gd,Yb,Er-PEG/PEI-FA nanoprobes are safe and reliable, and have potential applications in the field of tumor targeted multimodal imaging.

[Effect of oxidative stress on periprosthetic osteolysis induced by TCP wear particles in mouse calvaria and its mechanism].

OBJECTIVE: To explore the effect of oxidative stress on periprosthetic osteolysis induced by TCP wear particles in mouse calvaria and its mechanism. METHODS: Thirty-six male ICR mice were randomly divided into three groups (n=12):sham group, TCP wear particles (TCP) group and N-acetyl-L-cysteine (NAC) group. Aperiprosthetic osteolysis model in mouse was established by implanting 30 mg of TCP wear particles onto the surface of bilateral parietal bones following removal of the periosteum. On the 2nd day post-operation, NAC (1.0 mg/kg) was locally injected to the calvarium under the periosteum every other day for 2 weeks. Then, all the mice were sacrificed to obtain blood and the calvaria. Periprosthetic osteolysis in the mouse calvaria was observed by tartrate resistant acid phosphatase (TRAP) staining; serum levels of tumor necrosis factor-alpha (TNF-α), interleukin-1beta (IL-1ß), interleukin-6 (IL-6); total anti-oxidation capacity (T-AOC) and superoxide dismutase (SOD) activity were examined by ELISA and chemical colorimetry, respectively; protein levels of glucose-regulated protein 78 (GRP78), protein kinase R-like ER kinase (PERK), phospho-PERK (p-PERK), eukaryotic initiation factor 2α (eIF2α) and phospho-eIF2α (p-eIF2α) in periprosthetic bone tissue were detected by Western blot. RESULTS: Compared with sham group, serum levels of TNF-α, IL-1ß and IL-6, and osteolysis area were increased obviously in TCP group (P<0.05), and serum level of T-AOC and SOD activity were decreased significantly in TCP group (P<0.05), GRP78 expression, the ratio of p-PERK and PERK, p-eIF2α and eIF2α in the mouse calvaria of TCP group were up-regulated markedly. Compared with TCP group, serum levels of TNF-α, IL-1ß and IL-6, and osteolysis area were decreased markedly in NAC group (P<0.05), serum level of T-AOC and SOD activity were increased obviously in NAC group (P<0.05), and GRP78 expression, the ratio of p-PERK/PERK and p-eIF2α/eIF2α were obviously down-regulated. CONCLUSIONS: Inhibition of oxidative stress can prevent periprosthetic osteolysis induced by TCP wear particles, which may be mediated by inactivation of PERK/eIF2α signaling pathway.