Tumors are serious diseases threatening human health,and the early diagnosis is essential to improve treatment success and patient survival.The study of tumor gene expression data has become a major tool for revealing tumor disease mechanisms,in which artificial intelligence plays an important role.The potential advantages of supervised learning,unsupervised learning and deep learning in tumor prediction and classification are explored from the perspective of machine learning methods.Special attention is paid to the impact of feature selection algorithms on gene screening and their importance in high-dimensional gene expression data.By providing a comprehensive overview of the application and development of artificial intelligence in the analysis of tumor gene expression data,the study aims to provide an outlook for future research directions and promote further development.

Objective To verify the genetic characteristics associated with gastric cancer,and to propose a hybrid feature selection method for identifying target genes,further analyzing their significance and establishing a new diagnostic prediction model.Methods Analysis of variance in bioinformatics was performed on the original gastric cancer data,and then machine learning methods such as random forest,recursive feature elimination of support vector machine,and LASSO algorithm were used to screen gastric cancer associated genes,and the intersection of results was taken as the key gene set.The key genes were identified and verified through enrichment analysis.The diagnosis and prediction models based on 8 kinds of machine learning classification algorithms such as multi-layer perceptron,logistic regression and decision tree,were constructed using the key genes.Results The key genes selected by the hybrid feature selection method were closely related to the tumorigenesis and development.Eight key genes(TXNDC5,BMP8A,ONECUT2,COL10A1,JCHAIN,INHBA,LCTL and TRIM59)were identified as potential markers of good diagnostic efficacy in gastric cancer.The ROC curve and accuracy results demonstrated that among the 8 classification models,MLP is the best gastric cancer prediction model,with an accuracy of 97.77%,which was 3.83%higher than that of Xgboost gastric cancer prediction model.Conclusion The study identifies 8 key genes for the diagnosis and prevention of gastric cancer,and establishes the optimal prognosis model.

Objective: To evaluate the dietary quality of residents in Wenzhou City, Zhejiang Province, so as to provide the basis for future health education and nutrition intervention programs. Methods: A stratified multi-stage random sampling method was used to select residents aged 18 years and older in 6 counties (cities, districts) of Wenzhou City as the study subjects, “24-hour dietary review for 3 consecutive days” was adopted to collect dietary intake, and the diet balance index (DBI_16) scoring method was applied to evaluate the dietary quality. Results: This study analyzed the dietary quality of 406 residents in Wenzhou City, including 197 males (48.52%) and 209 females (51.48%). The majority of the residents were aged 18-44 years (254 residents, 62.56%). The median DBI total score was -31 (interquartile range, 8), and 404 residents had insufficient dietary intake, accounting for 99.51%. The median DBI positive score was 5 (interquartile range, 6), and 288 residents had appropriate dietary intake, accounting for 70.94%. The median DBI negative score was 37 (interquartile range, 6), and 210 residents had a high level of insufficient dietary intake, accounting for 51.72%. Five dietary patterns, namely A, B, C, E and F, were identified, with pattern B being the most dominant, accounting for 75.62% of the total (307 individuals). Patterns D, H, I and G were not observed. Conclusions The dietary quality of the residents surveyed indicates the existence of dietary imbalances, mainly manifesting as inadequate intake. It is recommended to strengthen nutritional and health guidance.

Objective:To screen ferroptosis genes related to prognosis of malignant pleural mesothelioma(MPM),explore the relationship between ferroptosis and tumor immune microenvironment and provide a new perspective for targeting and immunotherapy of MPM patients.Methods:The differentially expressed genes(DEGs)in MPM tumor group and normal group were analyzed in GEO database;intersection of DEGs and ferroptosis genes to obtain differentially expressed ferroptosis-related genes(DE-FRGs).GO,KEGG function enrichment and protein protein interaction(PPI)were used to identify the signal pathways mainly involved by DE-FRGs.The prognosis related ferroptosis genes were identified by univariate COX analysis.LASSO regression analysis was used to screen the best DE-FRGs for establishing the risk prediction model,and a risk prognosis model based on the best DE-FRGs was estab-lished by multivariate cox analysis to verify the prediction effect of the model.Finally,CIBERSORT and other algorithms were used to analyze tumor immune cell infiltration and evaluate immune microenvironment.Results:Twenty-four prognosis related DE-FRGs were screened,which were mainly concentrated in ferroptosis,transcriptional regulation and response to inorganic substances.A MPM risk prediction model based on five ferroptosis-related genes(ALDH3A2,CAV1,HRAS,CDCA3 and RRM2)was established and vali-dated.In the model,the proportion of CD8+T cells and macrophages in high-risk group were higher,while the proportion of B lympho-cytes was lower.In addition,PD-1,CTLA-4 and their ligands at immune checkpoint had higher expression status in high-risk group.Conclusion:The MPM risk prediction model based on five ferroptosis-related genes is established,and the immune status in the model is clarified.It provides a certain research basis for targeting and immunotherapy of MPM.The predictive ability of this model in MPM needs to be further verified in clinical practice to better predict disease stratification and treatment management.

Objective To investigate the predictive factors for death within 30 days after admission in patients with decompensated liver cirrhosis and acute kidney injury(AKI),and to establish and validate a nomogram prediction model.Methods The Joint Medical Record Management System of The First Affiliated Hospital of Nanchang University was used to obtain the patients with decompensated liver cirrhosis who were hospitalized in Department of Gastroenterology and Department of Infectious Diseases from January 2015 to December 2020,among whom 330 patients who met the 2015 International Club of Ascites diagnostic criteria for AKI were enrolled and divided into training group with 193 patients and validation group with 137 patients.A Cox regression analysis was used to investigate the predictive factors for death,and then a nomogram prediction model for the risk of death within 30 days after admission was established and validated.The independent-samples t-test was used for comparison of normally distributed continuous data between two groups,and a one-way analysis of variance was used for comparison between multiple groups,while the least significant difference t-test was used for further comparison between two groups;The Mann-Whitney U test was used for comparison of non-normally distributed continuous data between two groups,while the Kruskal-Wallis H test was used for comparison between multiple groups.The chi-square test or the Fisher's exact test was used for comparison of categorical data between groups.Results The prevalence rate of AKI was 16.5%in patients with decompensated liver cirrhosis.The 330 patients included in the study had a mean age of 53.6±12.4 years,and male patients accounted for 79.1%.The mortality rate was 50.0%within 30 days after admission,with a mortality rate of 46.6%in the training group and 54.7%in the validation group.The presence of acute-on-chronic liver failure(ACLF)on admission was an independent risk factor for the progression of AKI into stage 1(odds ratio=2.571,95%confidence interval:1.143-5.780,P=0.022).The nomogram based on white blood cell count,international normalized ratio,presence or absence of hepatic encephalopathy,and AKI stage on admission could well predict the risk of death with 30 days after admission,with a C-index of 0.680 in the training group and 0.683 in the validation group,and it was not inferior to CTP score and MELD score.Conclusion ACLF is an independent risk factor for the progression of AKI into stage 1.The nomogram prediction model established in this study can effectively predict the risk of death within 30 days after admission and thus has important guiding significance for the early identification and management of patients with decompensated liver cirrhosis and AKI.

Objective To observe the temporary loss(re-falling)of knee range of motion(ROM)during rehabilitation after arthroly-sis for post-traumatic knee stiffness,and analyze the factors related to it. Methods From July,2016 to September,2019,64 patients(68 knees)in Beijing Bo'ai Hospital accepted minimally inva-sive arthrolysis for post-traumatic knee stiffness,and were followed up for twelve months.ROM of flexion and extension of knee was measured before operation,and one,six and twelve weeks,and six and twelve months af-ter operation;while they were asssessed with Hospital for Special Surgery Knee Score(HSS).Multivariate Logis-tic regression was performed on re-falling. Results The ROM and HSS score improved as a whole after operation;however,HSS score improved constantly,but ROM decreased six weeks after operation compared with that one weeks after operation,involving 46 knees of 44 cases.Arthrolysis longer than 12 months from primary injuries,multiple complicated fracture and history of infection were the risk factors for re-falling(OR＞8.058,P＜0.05). Conclusion Minimally invasive arthrolysis is effective on knee function after arthrolysis for post-traumatic knee stiff-ness.However,re-falling of ROM may happen during rehabilitation.Delay of operation,multiple complicated fracture and history of infection may increase the risk of re-falling.

ObjectiveTo observe the clinical efficacy of Danzhi Jiangtang capsules with the functions of replenishing Qi， nourishing Yin， and dredging collaterals on patients with type 2 diabetes mellitus （T2DM） combined with lower-extremity macroangiopathy and serum levels of homocysteine （Hcy） and cystatin C （Cys C）. MethodA total of 80 eligible patients who were treated in the department of endocrinology of the First Affiliated Hospital of Anhui University of Chinese Medicine from May 2017 to December 2019 were randomized into the treatment group （40 cases） and control group （40 cases）. Both groups received the basic therapies for diabetes and Danzhi Jiangtang capsules （oral） was added to the treatment group. The levels of glycosylated hemoglobin A1c （HbA1c）， fasting plasma glucose （FPG）， 2-hour postprandial glucose （2 hPG）， fasting C-peptide （C-P）， and 2-hour postprandial C-peptide （2 hC-P）， triglycerides （TG）， total cholesterol （TC）， high density lipoprotein cholesterol （HDL-C）， and low density lipoprotein cholesterol （LDL-C） and serum levels of Hcy and Cys C were measured and the traditional Chinese medicine（TCM） syndrome scores were calculated before and after treatment in the two groups. ResultAfter 12 weeks of treatment， levels of HbA1c， FPG， and 2 hPG were lower than those before treatment in the two groups （P<0.05）. Levels of C-P （P<0.05） and 2 hC-P were higher than those before treatment in the two groups. After treatment， levels of HbA1c， FPG， and 2 hPG in the treatment group were lower than those in the control group （P<0.05）， while levels of C-P and 2 hC-P showed no significant difference between two groups. After treatment， the levels of TG， TC， and LDL-C were lower than those before treatment （P<0.05） and HDL-C level was higher than that before treatment （P<0.05） in both groups. After treatment， levels of TG and LDL-C in the treatment group were lower than those in the control group （P<0.05）， and levels of TC and HDL-C demonstrated no significant difference between two groups. After the treatment， the TCM syndrome score was lower than that before the treatment in both groups （P<0.05） and lower in the treatment group than in the control group （P<0.05）. The overall effective rate of the treatment group was higher than that of the control group （χ^２=7.585， P<0.05）. The levels of Cys C and Hcy were lower than those before treatment in the two groups （P<0.05） and lower in the treatment group than in the control group （P<0.05）. Doppler echocardiography of the lower limbs showed no obvious improvement in the control group after treatment. However， for the treatment group， slight decrease in intima-media thickness of the lower limb arteries and a slight reduction in the plaque area were observed， but the difference was not statistically significant. ConclusionDanzhi Jiangtang capsules has definite therapeutic effect on T2DM combined with lower-extremity macroangiopathy， which can improve glucolipid metabolism and reduce serum levels of Hcy and Cys C. This study can serve a reference for the prevention and treatment of T2DM combined with macroangiopathy.

Objective:To observe the clinical efficacy of Qigui Tangtongning Granules in the treatment of diabetic peripheral neuropathy (DPN) with qi deficiency and blood stasis.Methods:Prospective cohort study. A total of 80 DPN patients with Qi deficiency and blood stasis in Endocrinology Department of the First Affiliated Hospital of Anhui University of Chinese Medicine from May 2021 to May 2022 who met the inclusion criteria were divided into 2 groups by random number table method, with 40 cases in each group. The control group was treated with epalrestat on the basis of routine hypoglycemia, and the treatment group was treated with Qigui Tangtongning Granules on the basis of control group. Both groups were treated for 8 weeks. TCM syndromes were scored before and after treatment. Disease severity was assessed using the Toronto Clinical Scoring System (TCSS). The motor nerve conduction velocity (MNCV) and sensory nerve conduction velocity (SNCV) of median nerve and common peroneal nerve were detected by electromyography/induced potentiometer. Serum CRP, TNF-α and IL-6 were detected by ELISA, fasting blood glucose (FPG) and two hours post-meal blood glucose (2 hPG) were detected by automatic biochemical analyzer, and glycosylated hemoglobin (HbA1c) was detected by automatic HBA1C analyzer. Adverse reactions were recorded and clinical efficacy was evaluated.Results:The total effective rate was 95.0% (38/40) in the treatment group and 77.5% (31/40) in the control group, the difference between the two groups was statistically significant ( χ2=5.17, P=0.023). After treatment, the TCM syndrome score and TCSS score of the treatment group were significantly lower than those in the control group ( t=-3.19 and -7.63, P<0.01); Median nerve SNCV [(47.90±4.51) m/s vs. (44.76±3.72) m/s, t=3.40], MNCV [(53.79±3.65) m/s vs. (51.32±4.25) m/s, t=2.79] and common peroneal nerve SNCV [(44.21±2.08) m/s vs. (40.51±2.49) m/s, t=7.23], MNCV [(44.63±4.72) m/s vs. (41.36±4.87) m/s, t=3.05] were significantly higher than those in the control group ( P<0.01); FPG [(5.05±0.63) mmol/L vs. (7.05±1.23) mmol/L, t=-9.17], 2 hPG [(9.10±1.64) mmol/L vs. (12.19±2.61) mmol/L, t=-6.35], HbA1c [(6.79±0.90) % vs. (7.22±1.02) %, t=-2.02] were significantly lower than those in the control group ( P<0.01 or P<0.05); TNF-α [(15.75±5.44) ng/L vs. (32.01±5.33) ng/L, t=-13.51], hs-CRP [(2.58±0.80) mg/L vs. (3.79±1.04) mg/L, t=-5.83], IL-6 [(18.20±4.92) ng/L vs. (29.97±5.18) ng/L, t=-10.41] were significantly lower than those in the control group ( P<0.01). No obvious adverse reactions were observed in 2 groups during treatment. Conclusion:Qigui Tangtongning Granules combined with conventional Western medicine can improve nerve conduction velocity, reduce inflammation and improve clinical efficacy in DPN patients with Qi-deficiency and blood-stasis syndrome.

Objective:To retrospectively analyze the efficacy of video-assisted thoracoscopic surgery (VATS) in the treatment of solitary pulmonary nodule (SPN) and related factors affecting the diagnosis of benign and malignant SPN.Methods:The clinical data and surgical conditions of 317 patients with SPN who were resected by VATS and diagnosed by pathology in Chaoyang Central Hospital from January 2017 to December 2019 were retrospectively analyzed.The patient′s surgical efficacy and corresponding imaging features were explored.Results:Of the 317 patients, 124 (39.1%) cases underwent uniportal VATS and 193(60.9%) cases underwent multi-portal-portal VATS, 195(61.5%) cases had lobectomy, 16(5.1%) cases had anatomical segmentectomy, and 106(33.4%) cases had wedge resection. The operation time of the whole group was 50 - 210 (146.5 ± 34.7) min, intraoperative blood loss was 10 - 180 (69.3 ± 31.5) ml, thoracic drainage tube retention time was 2 - 7 (3.7 ± 1.4) d, postoperative hospital stay was 3 - 21 (6.4 ± 1.8) d, and the complication rate was 1.0%(3/317). No perioperative death occurred. Among all the nodules, 98(31.0%) cases were benign nodules and 219(69.1%) cases were malignant. Multivariate analysis showed that maximum nodule diameter, lobulated sign, burr sign, vascular bundle sign and pleural sag sign were independent related factors for the diagnosis of benign and malignant nodules ( P<0.05). Conclusions:VATS has less trauma and fewer complications, which has important clinical value for the diagnosis and treatment of benign and malignant SPN. Nodule maximum diameter, lobulated sign, burr sign, vascular bundle sign, pleural pit sign are independent related factors that affect the diagnosis of benign and malignant SPN.

Objective : To understand the spatial distribution of iodine in drinking water in Wenzhou,and to provide reference for preventing iodine deficiency disorders. Methods : Drinking water was sampled from 182 townships of all 11 counties under the jurisdiction of Wenzhou according to different ways of water supply. The iodine in water was detected by cerium sulfate catalytic spectrophotometry. The water iodine data was matched with the electronic map by ArcGIS10.2 to construct a spatial database; spatial autocorrelation analysis by GeoDa1.10 and spatial scanning analysis by SaTScan 9.4 were conducted to obtain the water iodine concentration range in Wenzhou. Results : The contents of iodine in 998 out of 1 008 drinking water samples were less than 10 μg/L,accounting for 99.01%. The median of water iodine in all townships of Wenzhou was 1.8 μg/L. The results of geospatial distribution analysis demonstrated that the iodine distribution in drinking water had positive spatial autocorrelation in Wenzhou（Moran's I= 0.40,Z=15.65,P< 0.05）; there were four kinds of local aggregation models for water iodine in 78 townships（P< 0.01）. Three cluster areas of the water iodine were detected by space scanning,with three townships in Dongtou as the first high cluster areas,seven coastal townships in Cangnan as the second high cluster areas and 49 mountainous townships in Yongjia,Yueqing and Lucheng as the low cluster areas. Conclusion The iodine in drinking water in Wenzhou was low and exists spatial aggregation.