With the continuous deepening of the concept of precision diagnosis and treatment for lung cancer, how to achieve higher efficiency and accuracy in the screening, diagnosis, and treatment pathways in clinical practice has become an important issue that urgently needs to be overcome. The current clinical difficulty lies in the fact that despite continuous advancements in imaging and molecular diagnostic technologies, there are still limitations in manual efficiency and subjective experience when it comes to massive data analysis and multi-scale feature extraction. Artificial intelligence (AI), especially algorithm systems based on deep learning, is an innovative technology capable of deeply empowering medical big data. This method utilizes algorithms such as convolutional neural networks, combined with radiomics, pathomics, and multi-modal data fusion analysis, demonstrating immense potential in early precise detection and benign-malignant differentiation of pulmonary nodules, digital pathological subtype recognition and non-invasive prediction of driver genes, precise 3D surgical planning and automatic delineation of radiotherapy target volumes, as well as dynamic risk warning during follow-up. This innovative technology provides a brand-new solution for realizing intelligent and individualized lung cancer diagnosis and treatment models. This consensus, based on the latest evidence from evidence-based medicine and combined with the development trends in the AI field and real-world clinical needs, was ultimately formed by gathering the consensus opinions of multidisciplinary experts in radiology, pathology, thoracic surgery, and other fields. The main content covers the application specifications of AI in the three core scenarios of lung cancer screening, diagnosis, and treatment, the technical standards for data collection and algorithm validation, as well as the ethical and regulatory challenges faced at the current stage. It aims to clarify the applicable boundaries of AI as a clinical auxiliary decision support tool, providing scientific guidance and standardized exploration directions for peers currently engaged in or planning to carry out AI-assisted clinical diagnosis, treatment, and translation of lung cancer.

OBJECTIVES: To develop a heterogeneous graph prediction method based on the fusion of multi-layer semantics and topological information for addressing the challenges in drug-target interaction prediction, including insufficient modeling of high-order semantic dependencies, lack of adaptive fusion of semantic paths, and over-smoothing of node features. METHODS: A heterogeneous graph network with multiple types of entities such as drugs, proteins, side effects, and diseases was constructed, and graph embedding techniques were used to obtain low-dimensional feature representations. An adaptive metapath search module was introduced to automatically discover semantic path combinations for guiding the propagation of high-order semantic information. A semantic aggregation mechanism integrating multi-head attention was designed to automatically learn the importance of each semantic path based on contextual information and achieve differentiated aggregation and dynamic fusion among paths. A structure-aware gated graph convolutional module was then incorporated to regulate the feature propagation intensity for suppressing redundant information and redcuing over-smoothing. Finally, the potential interactions between drugs and targets were predicted through an inner product operation. RESULTS: Compared with existing drug-target interaction prediction methods, the proposed method achieved an average improvement of 3.4% and 2.4%, 3.0% and 3.8% in terms of the area under the receiver operating characteristic curve (AUC) and the area under the precision-recall curve (AUPRC) on public datasets, respectively. CONCLUSIONS The drug-target interaction prediction method developed in this study can effectively extract complex high-order semantic and topological information from heterogeneous biological networks, thereby improving the accuracy and stability of drug-target interaction prediction. This method provides technical support and theoretical foundation for precise drug target discovery and targeted treatment of complex diseases.
Semantics ; Humans ; Drug Interactions ; Neural Networks, Computer ; Algorithms

Objective : To investigate the role of enhancer of zeste homolog 2(EZH2)-trimethylated lysine 27 of histone H3(H3K27me3) axis in the dedifferentiation of thyroid cancer and its clinical value as a potential target for the treatment of anaplastic thyroid cancer(ATC). Methods : Immunohistochemical SP method was used to detect the expression of EZH2, H3K27me3, paired box gene 8(PAX8), thyroglobulin(TG) and thyroid transcription factor 1(TTF1) in ATC and papillary thyroid carcinoma(PTC) and their adjacent tissues. The relationship between EZH2 and thyroid differentiation markers(PAX8, TTF1, TG) was further analyzed by gene expression omnibus(GEO) database. ATC cell lines 8305C and BHT-101 were culturedin vitro. Real-time reverse transcription PCR(RT-qPCR) was used to detect the expression of thyroid differentiation markers(TTF1, PAX8) mRNA in ATC cell lines treated with EZH2 inhibitor(GSK126), and evaluate the potential therapeutic effect of GSK126in vitro. The effects of GSK126 and BRAF inhibitor vemurafenib on the proliferation of ATC cell lines were observed by cell proliferation assay. Results : The expression of EZH2 in ATC tissues was significantly higher than that in papillary thyroid carcinoma and adjacent tissues(P<0.05). The expression of H3K37me3 in ATC tissues was significantly lower than that in PTC tissues(P<0.05). EZH2 was negatively correlated with PAX8 and TG expression levels, but not with TTF1 expression level.In vitroexperiments, GSK126 could reverse the expression of thyroid differentiation markers PAX8 and TTF1 in ATC cell lines. GSK126 combined with BRAF inhibitor vemurafenib could significantly inhibit the growth of ATC cell lines. Conclusion The EZH2-H3K27me3 axis plays an important role in regulating thyroid specific markers, and the inhibition of EZH2 by small molecular compounds is a promising target for ATC treatment in the future.

[Objective] To compare the diagnostic accuracy of three imaging modalities, inlducing CT urography (CTU), contrast-enhanced MRI (CE-MRI), and contrast-enhanced ultrasound (CEUS) in the qualitative diagnosis of renal space-occupying lesions. [Methods] A retrospective analysis was performed on 542 patients with renal lesions confirmed by surgical pathology in our hospital during Jan.2019 and May 2024.The diagnostic results of CTU, CE-MRI and CEUS were compared and analyzed based on the patients' clinical and pathological data. [Results] The diagnostic accuracy rate of CTU, CE-MRI and CEUS were 84.50%, 83.14% and 86.14%, respectively.For the 161 patients who underwent all three examinations, CEUS was significantly more accurate than CTU (84.16% vs. 77.02%, P=0.018), while there was no significant difference between CTU or CEUS and CE-MRI (79.81%) (P>0.05). Further analysis found that for lesions ≤4 cm, the accuracy of the three examinations was as follows: CEUS=CTU 79.55%, CE-MRI 76.14%, with no significant difference (P>0.05). However, for lesions >4 cm, CEUS ranked the first, followed by CE-MRI and CTU (89.73% vs. 84.25% vs. 73.97%), and CEUS and CE-MRI were better than CTU (P<0.05). Additionally, for the diagnosis of clear cell renal carcinoma and benign renal space-occupying lesions, there was no statistically significant difference among the three imaging modalities (P>0.05), while for the qualitative diagnosis of non-clear cell renal carcinoma, CEUS ranked the first, followed by CE-MRI and CTU (83.87% vs. 74.19% vs. 56.45%), and CE-MRI and CEUS were better than CTU (P<0.05). [Conclusion] All of them have important diagnostic value, and the appropriate selection should be based on patients' specifc conditions.CEUS and CE-MRI are more accurate in the qualitative diagnosis of renal space-occupying lesions than CTU, especially for large lesions and non-clear cell carcinoma.

Objective To investigate the effect of acute cold stimulation on the staufen1-mediated mRNA decay(SMD)pathway in brown adipose tissue of mice and the downstream regulated target genes.Methods Mice were subjected to acute cold stimulation(CS)at 4℃.After 48 hours,the brown adipose tissue of mice was extracted to detect the expression of genes including as Stau1,Ucp1 and Pparγ,and compared with mice in room temperature control group(RT).Transcriptomic sequencing was performed on the brown adipose tissue of mice in the CS group and in the RT group,and the functional enrichment analysis of differential genes was performed on the sequencing re-sults.The Stau1 gene was knocked out in the brown adipocytes of mice using CRISPR-Cas9 technology,and the ex-pression of thermogenic genes after knockout was analyzed.Results Acute cold stimulation induced the expression of Stau1 gene and promoted the degradation of downstream target genes Serpineb1,Klf2 and c-Jun in the SMD pathway(P＜0.05).After Stau1 knockout,the glycolipid metabolism pathway of brown adipocytes in mice was significantly up-regulated,and the expression of thermogenesis-related genes Ucp1,Prdm16,ATP5o,Dio2 and Pgc1α was up-regulated(P＜0.05).Conclusions Acute cold stimulation promotes the SMD pathway in brown adipose tissue of mice,and SMD pathway mainly regulates the metabolic and thermogenic pathways in brown adipocytes.

Objective:This study aims to evaluate the performance of digital PCR (dPCR) detecting multiple and single copies genes of the Epstein-Barr virus (EBV) for nucleic acid quantification and explore their applicability in clinical settings.Methods:Compared the sensitivity, specificity, precision, lower limit of detection (LoD), and linearity for multicopy BamHI-W dPCR and single-copy EBNA1 dPCR systems. Linear regression analysis using the least squares method was employed to evaluate the linearity. Additionally, we analyzed plasma samples from 182 patients with suspected EBV-related diseases between January and July 2022 at the Southern Medical University Southern Hospital, using both dPCR and quantitative PCR (qPCR) for EBV DNA quantification. Linear regression analysis using the least squares method was conducted to assess their quantitative correlation.Results:The dPCR systems for both multicopy and single-copy genes showed excellent linearity ( R 2 values of 0.992 and 0.997, respectively, both P<0.001). The LoD were 188 IU/ml for BamHI-W gene and 358 IU/ml for EBNA1 gene dPCR systems. The logarithmic coefficient of variation ( CV) values for high-concentration samples (1 000 000 IU/ml) were 0.34% and 0.21% for the BamHI-W gene and EBNA1 gene dPCR assays, respectively, while for low-concentration samples (5 000 IU/ml) were 0.98% and 0.64%, respectively. In the detection of seven common clinical infectious pathogens and EBV positive samples, only EBV-positive samples yielded positive signals in the dPCR detection system, with no cross-reaction with other pathogens. In 182 samples, the positive detection rates were 47.80% (87/182) for BamHI-W gene and 35.16% (64/182) for EBNA1 gene dPCR, compared to 43.41% (79/182) for qPCR. Linear correlation analysis with qPCR showed R2 values of 0.837 for BamHI-W gene and 0.763 for EBNA1 gene dPCR (both P<0.001). The BamHI-W gene copy number ranged from 3 to 18 copies per clinical sample, with patient-specific variations. There was a high consistency in viral load trends between the multicopy BamHI-W gene and single-copy EBNA1 gene dPCR systems within individual patients. Conclusions:The dPCR methods detecting EBV multiple and single copies genes showed high sensitivity, specificity, precision, and quantitative accuracy, suitable for clinical sample analysis. The multicopy BamHI-W gene dPCR method notably enhances detection sensitivity and can be used as a supplement to current EBV DNA load detection methods, especially in low-concentration samples. For within-patient EBV DNA monitoring, the multicopy gene method proves more effective, while inter-patient comparisons might necessitate single-copy gene methods or normalize them using the same standard.

Objective:To analyze the status of respiratory pathogen detection and the clinical features in children with Mycoplasma pneumoniae pneumonia (MPP). Methods:A prospective, multicenter study was conducted to collect clinical data, including medical history, laboratory examinations and multiplex PCR tests of children diagnosed with MPP from 4 hospitals in China between November 15 th and December 20 th, 2023. The multiplex PCR results and clinical characteristics of MPP children in different regions were analyzed. The children were divided into severe and mild groups according to the severity of the disease. Patients in the severe group were further divided into Mycoplasma pneumoniae (MP) alone and Multi-pathogen co-detection groups based on whether other pathogens were detected besides MP, to analyze the influence of respiratory pathogen co-detection rate on the severity of the disease. Mann-Whitney rank sum test and Chi-square test were used to compare data between independent groups. Results:A total of 298 children, 136 males and 162 females, were enrolled in this study, including 204 children in the severe group with an onset age of 7.0 (6.0, 8.0) years, and 94 children in the mild group with an onset age of 6.5 (4.0, 7.8) years. The level of C-reactive protein, D-dimer, lactic dehydrogenase (LDH) were significantly higher (10.0 (5.0, 18.0) vs. 5.0 (5.0, 7.5) mg/L, 0.6 (0.4, 1.1) vs. 0.5 (0.3, 0.6) mg/L, 337 (286, 431) vs. 314 (271, 393) U/L, Z=2.02, 2.50, 3.05, all P<0.05), and the length of hospitalization was significantly longer in the severe group compared with those in mild group (6.0 (6.0, 7.0) vs. 5.0 (4.0, 6.0) d, Z=4.37, P<0.05). The time from onset to admission in severe MPP children was significantly shorter than that in mild MPP children (6.0 (5.0, 9.5) vs. 9.0 (7.0, 13.0) d, Z=2.23, P=0.026). All patients completed the multiplex PCR test, with 142 cases (47.7%) MPP children detected with 21 pathogens including adenovirus 25 cases (8.4%), human coronavirus 23 cases (7.7%), rhinovirus 21 cases (7.0%), Streptococcus pneumoniae 21 cases (7.0%), influenza A virus 18 cases (6.0%). The pathogens with the highest detection rates in Tianjin, Shanghai, Wenzhou and Chengdu were Staphylococcus aureus at 10.7% (8/75), adenovirus at 13.0% (10/77), adenovirus at 15.3% (9/59), and both rhinovirus and Haemophilus influenzae at 11.5% (10/87) each. The multi-pathogen co-detection rate in severe MPP children was significantly higher than that in mild MPP group (52.9% (108/204) vs. 36.2% (34/94), χ2=10.62, P=0.005). Among severe MPP children, there are 89 cases in the multi-pathogen co-detection group and 73 cases in the simple MPP group. The levels of LDH, D-dimer and neutrophil counts in the multi-pathogen co-detection group were significantly higher than those in the simple MPP group (348 (284, 422) vs. 307 (270, 358) U/L, 0.8 (0.5, 1.5) vs. 0.6 (0.4, 1.0) mg/L, 4.99 (3.66, 6.89)×10 9vs. 4.06 (2.91, 5.65)×10 9/L, Z=5.17, 4.99, 6.11, all P<0.05). Conclusions:The co-detection rate of respiratory pathogens, LDH and D-dimer in children with severe MPP were higher than those with mild MPP. Among severe MPP children the stress response of children in co-detection group was more serious than that of children with simple MPP.

Coronary restenosis is an important cause of poor long-term prognosis in patients with coronary heart disease. Here, we show that lysine methyltransferase SMYD2 expression in the nucleus is significantly elevated in serum- and PDGF-BB-induced vascular smooth muscle cells (VSMCs), and in tissues of carotid artery injury-induced neointimal hyperplasia. Smyd2 overexpression in VSMCs (Smyd2-vTg) facilitates, but treatment with its specific inhibitor LLY-507 or SMYD2 knockdown significantly inhibits VSMC phenotypic switching and carotid artery injury-induced neointima formation in mice. Transcriptome sequencing revealed that SMYD2 knockdown represses the expression of serum response factor (SRF) target genes and that SRF overexpression largely reverses the inhibitory effect of SMYD2 knockdown on VSMC proliferation. HDAC3 directly interacts with and deacetylates SRF, which enhances SRF transcriptional activity in VSMCs. Moreover, SMYD2 promotes HDAC3 expression via tri-methylation of H3K36 at its promoter. RGFP966, a specific inhibitor of HDAC3, not only counteracts the pro-proliferation effect of SMYD2 overexpression on VSMCs, but also inhibits carotid artery injury-induced neointima formation in mice. HDAC3 partially abolishes the inhibitory effect of SMYD2 knockdown on VSMC proliferation in a deacetylase activity-dependent manner. Our results reveal that the SMYD2-HDAC3-SRF axis constitutes a novel and critical epigenetic mechanism that regulates VSMC phenotypic switching and neointimal hyperplasia.

ObjectiveTo investigate the short-term knee function recovery of patients undergoing total knee arthroplasty (TKA) after discharge and analyze its related factors. MethodsFrom December, 2022 to April, 2023, 140 adult patients who underwent TKA in Nanjing Drum Tower Hospital and being about to be discharged were selected as the survey subjects using convenient sampling. Potential influencing factors were selected based on a literature review. They were investigated with general information questionnaire, Rehabilitation Exercise Compliance Scale, Self-Efficacy Scale for Rehabilitation Exercise (SER), Visual Analog Scale for pain (VAS), and joint range of motion measurements before discharge, and were investigated with the Hospital for Special Surgery Knee Score (HSS) one month after discharge. ResultsA total of 130 patients finished follow-up. One month after discharge, the HSS score ranged from 40 to 82, with an average of (70.89±6.26). Multiple linear regression analysis revealed that Body Mass Index (B = -0.423, 95%CI -0.622 to -0.224, P < 0.001), pre-discharge VAS (B = -1.016, 95%CI -1.198 to -0.113, P = 0.028), rehabilitation exercise compliance (B = 0.267, 95%CI 0.121 to 0.413, P < 0.001), SER (B = 0.478, 95%CI 0.315 to 0.642, P < 0.001), and knee joint flexion contracture angle (B = -0.251, 95%CI -0.414 to -0.088, P = 0.003) could influence HSS score one month after discharge (R² = 0.615, F =17.106, P < 0.001). ConclusionPatients after TKA have recovered well in short time after discharge, however, there is still significant room for improvement. Clinical healthcare providers should design and implement appropriate interventions based on related factors to improve the function.

The prevalence of periodontal disease in Chinese population is more than 90％.The present treatment techniques can only control the development of the disease,inducement of bone tissue regeneration is a promising strategy and a challenge for the treatment.Exosomes are multivesicle structures derived from endosomes.More and more studies have been conducted on their application in perio-dontal regeneration.This paper reviews the application of exosome in periodontal regeneration in recent years,which is expected to pro-vide new idea for periodontal regeneration therapy.