ObjectiveBased on traditional quality evaluation of Gardeniae Fructus（GF） recorded in historical materia medica， this study systematically compared the quality differences between wild and cultivated GF from morphological characteristics， microscopic features， and contents of primary and secondary metabolites. MethodsVernier calipers and analytical balances were used to measure the length， diameter and individual fruit weight of wild and cultivated GF， and the aspect ratio was calculated. A colorimeter was used to determine the chromaticity value of wild and cultivated GF， and the paraffin sections of them were prepared by safranin-fast green staining and examined under an optical microscope to observe their microstructure. Subsequently， the contents of water-soluble and alcohol-soluble extracts of wild and cultivated GF were detected by hot immersion method under the general rule 2201 in volume Ⅳ of the 2020 edition of the Pharmacopoeia of the People's Republic of China， the starch content was measured by anthrone colorimetric method， the content of total polysaccharides was determined by phenol-sulfuric acid colorimetric method， the sucrose content was determined by high performance liquid chromatography coupled with evaporative light scattering detection（HPLC-ELSD）， and the contents of representative components in them were measured by ultra-performance liquid chromatography（UPLC）. Finally， correlation analysis was conducted between quality traits and phenotypic traits， combined with multivariate statistical analysis methods such as principal component analysis（PCA） and orthogonal partial least squares-discriminant analysis（OPLS-DA）， key differential components between wild and cultivated GF were screened. ResultsIn terms of traits， the wild GF fruits were smaller， exhibiting reddish yellow or brownish red hues with significant variation between batches. While the cultivated GF fruits are larger， displaying deeper orange-red or brownish red. The diameter and individual fruit weight of cultivated GF were significantly greater than those of wild GF， while the blue-yellow value（b^*） of wild GF was significantly higher than that of cultivated GF. In the microstructure， the mesocarp of wild GF contained numerous scattered calcium oxalate cluster crystals， while the endocarp contained stone cell class round， polygonal or tangential prolongation， undeveloped seeds were visible within the fruit. In contrast， the mesocarp of cultivated GF contained few calcium oxalate cluster crystals， or some batches exhibited extremely numerous cluster crystals. The stone cells in the endocarp were predominantly round-like， with the innermost layer arranged in a grid pattern. Seeds were basically mature， and only a few immature seeds existed in some batches. Regarding primary metabolite content， wild GF exhibited significantly higher total polysaccharide level than cultivated GF（P<0.01）. In category-specific component content， wild GF exhibited significantly higher levels of total flavonoids and total polyphenols compared to cultivated GF（P<0.01）. Analysis of 12 secondary metabolites revealed that wild GF exhibited significantly higher levels of Shanzhiside， deacetyl asperulosidic acid methyl ester， gardenoside and chlorogenic acid compared to cultivated GF（P<0.01）. Conversely， the contents of genipin 1-gentiobioside， geniposide and genipin were significantly lower in wild GF（P<0.01）. ConclusionThere are significant differences between wild and cultivated GF in terms of traits， microstructure， and contents of primary and secondary metabolites. At present， the quality evaluation system of cultivated GF remains incomplete， and this study provides a reference for guiding the production of high-quality GF medicinal materials.

Objective To construct the training program for the digital and intelligent capabilities of specialized nurses in the Central Sterile Supply Department(CSSD),and conduct preliminary practice to provide talent support for the intelligent development of CSSD.Methods From February to April 2024,based on the core technologies of digital intelligence and related core capabilities,a training program for digital intelligence-related competencies of CSSD specialized nurses was constructed using literature review and the Delphi expert consultation method.From July to August 2024,the program was initially implemented in the training of CSSD specialized nurses.The nurses'information competency before and after the training was compared,and the nurses' satisfaction with the digital intelligence-related training program was assessed.Results This study conducted 2 rounds of expert consultation via questionnaire.The effective recovery rate of the questionnaires in both rounds was 100％.The expert authority coefficients were 0.790 and 0.800,respectively,and the variation coefficients ranged from 0 to 0.229 and 0 to 0.105.Ultimately,a training program for the digital-related competencies of CSSD specialty nurses was established,which includes 4 components:training objectives,training content,training methods,and assessment methods.Specifically,there were 3 indicators at the first level and 14 at the second level for training objectives,6 indicators at the first level and 32 at the second level for training content,and 6 indicators at the first level for training methods and assessment methods.After the implementation of the training program,the information competency of the nurses in all dimensions and the total score were significantly higher than those before training(P＜0.05).Moreover,the average scores for the training content,training methods,and assessment methods were all above 3 points,indicating a high overall satisfaction among the nurses.Conclusion The construction process of the training program for the digital and intelligent capabilities of CSSD specialty nurses is scientific and reliable.The content is highly practical and distinctive in its specialty.The training methods and assessment approaches are diverse.This program can enhance nurses' information competency and provide a reference for the implementation of digital and intelligent training for CSSD specialty nurses.

Objective:To investigate the prognostic factors for esophageal cancer after radical resection and the application value of machine learning prediction model.Methods:The retrospective cohort study was conducted. The clinicopatholigical data of 406 esophageal cancer patients who were admitted to Qilu Hospital of Shandong University from January 2018 to March 2022 were collected. There were 357 males and 49 females, aged (64±8)years. All patients underwent radical resection of esophageal cancer. The 406 patients were randomly divided into a training set of 285 cases and a validation set of 121 cases at a 7∶3 ratio based on a random number table. The training set was used to construct prediction model, and the validation set was used to validate prediction model. Patients were divided into high-risk group and low-risk group based on risk scores. Observation indicators: (1) follow-up of patients and analysis of influencing factors for prognosis; (2) construction and validation of machine learning prediction models. Comparison of measurement data with normal distribution between groups was conducted using the independent sample t test. Comparison of measurement data with skewed distribution between groups was conducted using the Mann-Whitney U test. Comparison of count data between groups was conducted using the chi-square test. Comparison of ordinal data between groups was conducted using the rank sum test. The Kaplan-Meier method was used to calculate survival rate and plot survival curve, and the Log-rank test was used for survival analysis. The Cox proportional hazard regression model was used for univariate and multivariate analyses. Independent influencing factors were included, and data processing, machine learning model construction, and visualization were performed using R packages including random survival forest (RSF), gradient boosting machine (GBM), least absolute shrinkage and selection operator Cox regression (LASSO-Cox), Cox proportional hazards model boosting (CoxBoost), survival support vector machine (survivalsvm), extreme gradient boosting (XGBoost), supervised principal component analysis (SuperPC), and Cox partial least squares regression (plsRcox). Receiver operating characteristic (ROC) curves were drawn, and sensitivity, specificity, and area under the curve (AUC) were calculated. The Delong test was used to assess the differences in AUC among different models in the training set, and the time-dependent ROC was used to compare the predictive performance of different models. Calibration curves were used to evaluate model accuracy, and decision curve analysis (DCA) was used to evaluate overall net benefit. Results:(1) Follow-up of patients and analysis of influencing factors for prognosis. All 406 patients were followed up postoperatively for 28(range, 6-36)months, with 1- and 3-year overall survival rate of 86.5% and 40.9%, respectively. The 285 patients in the training set were followed up postoperatively for 30(range, 6-36)months, with 1- and 3-year overall survival rate of 85.1% and 35.5%, respectively. The 121 patients in the validation set were followed up postoperatively for 25(range, 6-36)months, with 1- and 3-year overall survival rate of 87.0% and 43.2%, respectively. There was no significant difference in postoperative overall survival rate between the training set and the validation set ( χ2=3.20, P>0.05). Results of multivariate analysis showed that left thoracic surgical approach, preopera-tive neutrophil count, vascular invasion, perineural invasion, pathological T2-4 stage, pathological N2-3 stage, and postoperative pneumonia were independent risk factors affecting postoperative survival of 285 patients in the training set ( hazard ratio=1.466, 1.037, 1.482, 1.549, 5.268, 7.727, 22.202, 2.539, 2.686, 1.425, 95% confidence interval as 1.026-2.096, 1.003-1.073, 1.008-2.179, 1.105-2.170, 1.201-23.099, 1.833-32.576, 4.734-104.128, 1.577-4.087, 1.631-4.422, 1.018-1.994, P<0.05). (2) Construction and validation of machine learning prediction models. Independent risk factors affecting postoperative survival were included to construct RSF, GBM, LASSO-Cox, CoxBoost, survivalsvm, XGBoost, SuperPC, and plsRcox machine learning prediction models. Results of Delong test showed that there were significant differences in the AUC of RSF and GBM from the other six models ( P<0.05). Results of time-dependent ROC curve showed that all 8 machine learning predic-tion models had good discriminative ability in the training cohort, among which the RSF machine learning prediction model had the best predictive performance. Results of calibration curve showed that the RSF machine learning prediction model fitted well for predicting postoperative 1-, 2-, and 3-year overall survival in the training cohort, with high consistency with actual results. Results of decision curve analysis showed that within a threshold range of 0-0.80, the RSF machine learning prediction model provided a better overall net benefit. Further analysis showed that in the validation set, the AUC of RSF machine learning prediction model for postoperative 1-, 2-, and 3-year survival prediction were 0.786 (95% confidence interval as 0.609-0.962), 0.774 (95% confidence interval as 0.676-0.873), and 0.750 (95% confidence interval as 0.652-0.848), respectively. Results of calibration curve showed that the RSF machine learning prediction model fitted well for predicting postopera-tive 1-, 2-, and 3-year overall survival in the validation set, with high consistency with actual results. In the training set, the optimal cutoff value of the RSF machine learning prediction model risk score was 11.7. Patients with risk score ≥11.7 were classified as the high-risk group, and those with risk score <11.7 as the low-risk group. The median survival times of the two groups were 18.0 months and >36.0 months, respectively, showing a significant difference between them ( χ2=73.30, P<0.05). In the validation set, the optimal cutoff value of the RSF machine learning prediction model risk score was 11.7. Patients with risk score ≥11.7 were classified as the high-risk group, and those with risk score<11.7 as the low-risk group. The median survival times of the two groups were 17.0 months and>36.0 months for the high-risk and low-risk groups, respectively, showing a significant difference between them ( χ2=35.20, P<0.05). Conclusions:Left thoracic surgical approach, preoperative neutrophil count, vascular invasion, perineural invasion, pathological T2-4 stage, pathological N2-3 stage, and postoperative pneumonia are independent risk factors affecting survival of esophageal cancer patients after radical resection. The RSF machine learning prediction model constructed based on these factors can effectively distinguish the survival prognosis of high-risk and low-risk patients.

Objective:To explore the effect of ginsenoside Rg1 on spermatogenic dysfunction in mice caused by diabetes and its mechanism of action.Methods:Eighteen male C57BL mice were randomly divided into control group, the model group and the ginsenoside Rg1 group by completely random method, with 6 mice in each group. Type 2 diabetes models were established in the model group and the ginsenoside Rg1 group by a high-fat diet combined with intraperitoneal injection of streptozotocin, while control group was injected with the same amount of normal saline. After successful modeling, control group was given a regular diet for 8 weeks, while the model group and ginsenoside Rg1 group were given a high-fat diet for 8 weeks. The ginsenoside Rg1 group was also treated with ginsenoside Rg1 medication. Reproductive hormone levels were detected by enzyme-linked immunosorbent assay test kits, and Western blotting was used to detect the expressions of apoptosis-related proteins (Bcl2 protein, Caspase-3 protein, Bax protein), autophagy-related proteins (P62, LC3Ⅰ, LC3Ⅱ, Beclin1), β-Catenin protein, mTOR protein, LAMP1 protein and transcription factor EB. The body weight, blood glucose levels, testicular index of mice in each group were compared, as well as the testicular injury status.Results:The body weight [(18.77±1.14) g], testosterone level [(141.07±8.47) ng/L], follicle-stimulating hormone level [(9.19±0.74) U/L], and luteinizing hormone level [(1 497.91±99.57) pg/L] of mice in the model group were significantly lower than those in the control [(31.57±2.35) g, P<0.001; (171.50±11.76) ng/L, P<0.001; (12.46±1.54) U/L, P<0.001; (1 807.29±92.76) pg/L, P<0.001]; fasting blood glucose level [(20.82±1.11) mmol/L], glycosylated hemoglobin (12.67%±1.03%), the testis index (0.65%±0.03%) were significantly higher than those in the control [(6.40±1.34) mmol/L, P<0.001; 5.17%±1.17%, P<0.001; 0.48%±0.04%, P<0.001]. Compared with the model group, the body weight [(22.62±0.92) g, P=0.023], testosterone level [(172.63±9.20) ng/L, P<0.001], follicle-stimulating hormone level [(12.37±1.15) U/L, P<0.001], and luteinizing hormone level [(1 847.80±108.80) pg/L, P<0.001] of mice in the ginsenoside Rg1 group increased significantly, fasting blood glucose level [(18.63±1.14) mmol/L, P=0.017], glycosylated hemoglobin (8.50%±1.05%, P<0.001) and testicular index (0.54%±0.02%, P<0.001) decreased significantly. Compared with the control, the expressions of P62 ( P=0.039), LC3Ⅱ/LC3Ⅰ( P<0.001), Beclin1 ( P=0.002) and mTOR ( P=0.036) in the testicular tissue of mice in the model group all increased, the expression of β-Catenin ( P<0.001), LAMP1 ( P=0.005), transcription factor EB ( P<0.001) all decreased. Compared with the model group, the expressions of autophagy-related proteins P62 ( P=0.048), LC3Ⅱ/LC3Ⅰ( P<0.001) , Beclin1 ( P=0.023) and mTOR ( P=0.005) in the ginsenoside Rg1 group all decreased, while the expression of β-Catenin ( P=0.001), LAMP1 ( P=0.011) and transcription factor EB ( P=0.022) all increased. Transmission electron microscopy detected a decrease in the number of autophagosomes in the testicles of mice in the model group, and it improved after drug intervention. The HE staining showed that the testes of mice in the model group exhibited phenotypes such as the shedding and disorganization of spermatogenic cells, while ginsenoside Rg1 was able to improve these phenotypes. Conclusion:Ginsenoside Rg1 can improve testicular injury caused by diabetes in mice by regulating autophagy.

Chronic epididymitis (CE) is a long-standing inflammatory condition of the epididymis caused by unresolved acute infections, chronic infections, medication use, or other factors. Clinically, it is characterized by persistent dull pain or a dragging sensation in one or both sides of the scrotum. The disease course typically exceeds three months and is marked by insidious onset and recurrent episodes. Current studies suggest that CE may disrupt the epididymal microenvironment through multiple pathological processes, including local inflammatory responses, oxidative stress, fibrotic remodeling, and autophagy. These alterations impair sperm maturation, transport, and capacitation, thereby contributing to male reproductive dysfunction and infertility. This review summarizes the major etiologies and pathophysiological characteristics of CE and its impact on male reproductive function. It focuses on the roles of inflammatory cytokines and related signaling pathways, oxidative stress mechanisms, and fibrotic progression in the pathogenesis of CE. Moreover, it explores targeted therapeutic strategies based on these mechanisms, aiming to provide a theoretical basis for identifying key molecular targets and signaling pathways involved in CE-induced male reproductive impairment.

Objective To investigate the effects of different anesthesia depths on stress states and inflammatory mediators in patients undergoing video-assisted thoracoscopic lobectomy.Methods A total of 89 lung cancer patients who underwent video-assisted thoracoscopic lobectomy were selected as study subjects.Based on intraoperative bispectral index(BIS)range,the patients were divided into deep anesthesia group(BIS of 40 to＜50,n=45)and shallow anesthesia group(BIS of 50 to＜60,n=44).Vital signs(mean arterial pressure,heart rate and blood oxygen saturation),anesthesia re-covery time,extubation time,dosage of vasoactive drugs,postoperative pain intensity[Visual Ana-logue Scale(VAS)],postoperative analgesic dosage,perioperative stress state[prostaglandin E2(PGE2),nerve growth factor(NGF)and substance P(SP)],levels of inflammatory mediators[neuron-specific enolase(NSE),tumor necrosis factor-α(TNF-α)and S100β protein]at different time points(before anesthesia induction,immediately after intubation,before lesion resection and at the end of surgery)and the incidence of anesthesia-related adverse reactions were compared between the two groups.Results Before lesion resection and at the end of surgery,the mean arterial pressure and heart rate in the deep anesthesia group were significantly lower than those in the shallow anesthe-sia group(P＜0.05).The anesthesia recovery time and extubation time in the deep anesthesia group were significantly longer than those in the shallow anesthesia group(P＜0.05).At the end of surgery and on postoperative day one,the levels of PGE2,NGF and SP in the deep anesthesia group were significantly lower than those in the shallow anesthesia group,while the levels f NSE,TNF-α and S100β protein were significantly higher than those in the shallow anesthesia group(P＜0.05).There were no significant differences in the dosage of vasoactive drugs,VAS scores,sufentanil dos-age and the incidence of anesthesia-related adverse reactions between thetwo groups(P＞0.05).Conclusion During one-lung ventilation in patients undergoing video-assisted thoracoscopic surgery lobectomy,deep anesthesia can effectively control surgical stress and maintain stability of intraopera-tive hemodynamics,but it is associated with delayed postoperative awakening and more pronounced inflammatory response.Shallow anesthesia results in faster postoperative awakening and lower levels of inflammatory mediators,but it is associated with more significant intraoperative stress response and unstable hemodynamics.

Objective:To explore the effect of ginsenoside Rg1 on spermatogenic dysfunction in mice caused by diabetes and its mechanism of action.Methods:Eighteen male C57BL mice were randomly divided into control group, the model group and the ginsenoside Rg1 group by completely random method, with 6 mice in each group. Type 2 diabetes models were established in the model group and the ginsenoside Rg1 group by a high-fat diet combined with intraperitoneal injection of streptozotocin, while control group was injected with the same amount of normal saline. After successful modeling, control group was given a regular diet for 8 weeks, while the model group and ginsenoside Rg1 group were given a high-fat diet for 8 weeks. The ginsenoside Rg1 group was also treated with ginsenoside Rg1 medication. Reproductive hormone levels were detected by enzyme-linked immunosorbent assay test kits, and Western blotting was used to detect the expressions of apoptosis-related proteins (Bcl2 protein, Caspase-3 protein, Bax protein), autophagy-related proteins (P62, LC3Ⅰ, LC3Ⅱ, Beclin1), β-Catenin protein, mTOR protein, LAMP1 protein and transcription factor EB. The body weight, blood glucose levels, testicular index of mice in each group were compared, as well as the testicular injury status.Results:The body weight [(18.77±1.14) g], testosterone level [(141.07±8.47) ng/L], follicle-stimulating hormone level [(9.19±0.74) U/L], and luteinizing hormone level [(1 497.91±99.57) pg/L] of mice in the model group were significantly lower than those in the control [(31.57±2.35) g, P<0.001; (171.50±11.76) ng/L, P<0.001; (12.46±1.54) U/L, P<0.001; (1 807.29±92.76) pg/L, P<0.001]; fasting blood glucose level [(20.82±1.11) mmol/L], glycosylated hemoglobin (12.67%±1.03%), the testis index (0.65%±0.03%) were significantly higher than those in the control [(6.40±1.34) mmol/L, P<0.001; 5.17%±1.17%, P<0.001; 0.48%±0.04%, P<0.001]. Compared with the model group, the body weight [(22.62±0.92) g, P=0.023], testosterone level [(172.63±9.20) ng/L, P<0.001], follicle-stimulating hormone level [(12.37±1.15) U/L, P<0.001], and luteinizing hormone level [(1 847.80±108.80) pg/L, P<0.001] of mice in the ginsenoside Rg1 group increased significantly, fasting blood glucose level [(18.63±1.14) mmol/L, P=0.017], glycosylated hemoglobin (8.50%±1.05%, P<0.001) and testicular index (0.54%±0.02%, P<0.001) decreased significantly. Compared with the control, the expressions of P62 ( P=0.039), LC3Ⅱ/LC3Ⅰ( P<0.001), Beclin1 ( P=0.002) and mTOR ( P=0.036) in the testicular tissue of mice in the model group all increased, the expression of β-Catenin ( P<0.001), LAMP1 ( P=0.005), transcription factor EB ( P<0.001) all decreased. Compared with the model group, the expressions of autophagy-related proteins P62 ( P=0.048), LC3Ⅱ/LC3Ⅰ( P<0.001) , Beclin1 ( P=0.023) and mTOR ( P=0.005) in the ginsenoside Rg1 group all decreased, while the expression of β-Catenin ( P=0.001), LAMP1 ( P=0.011) and transcription factor EB ( P=0.022) all increased. Transmission electron microscopy detected a decrease in the number of autophagosomes in the testicles of mice in the model group, and it improved after drug intervention. The HE staining showed that the testes of mice in the model group exhibited phenotypes such as the shedding and disorganization of spermatogenic cells, while ginsenoside Rg1 was able to improve these phenotypes. Conclusion:Ginsenoside Rg1 can improve testicular injury caused by diabetes in mice by regulating autophagy.

Chronic epididymitis (CE) is a long-standing inflammatory condition of the epididymis caused by unresolved acute infections, chronic infections, medication use, or other factors. Clinically, it is characterized by persistent dull pain or a dragging sensation in one or both sides of the scrotum. The disease course typically exceeds three months and is marked by insidious onset and recurrent episodes. Current studies suggest that CE may disrupt the epididymal microenvironment through multiple pathological processes, including local inflammatory responses, oxidative stress, fibrotic remodeling, and autophagy. These alterations impair sperm maturation, transport, and capacitation, thereby contributing to male reproductive dysfunction and infertility. This review summarizes the major etiologies and pathophysiological characteristics of CE and its impact on male reproductive function. It focuses on the roles of inflammatory cytokines and related signaling pathways, oxidative stress mechanisms, and fibrotic progression in the pathogenesis of CE. Moreover, it explores targeted therapeutic strategies based on these mechanisms, aiming to provide a theoretical basis for identifying key molecular targets and signaling pathways involved in CE-induced male reproductive impairment.

Erectile dysfunction(ED)is a common sexual dysfunction in adult males.With the accelerated pace of modern life and lifestyle changes,the prevalence of ED and its associated comorbidities have been steadily rising.Problems such as premature ejaculation,hypertension,hyperlipidemia,diabetes,prostate diseases,and infertility all interact with and aggravate ED,thereby endangering the overall health of men.Phosphodiesterase type 5 inhibitors(PDE5i)is the first-line pharmacotherapy for ED.Tadalafil,currently the only long-acting PDE5i approved for clinical use,has received mar-keting authorization in China for the treatment of ED since 2005.In 2013,the once-a-day continuous regimen was intro-duced as a novel treatment paradigm.And the indication was expanded to ED coexisting with benign prostatic hyperplasia in 2019.Accumulating clinical experience and evidence-based data consistently demonstrate its efficacy and safety across ED and ED-related comorbidities.This review summarizes the pharmacological profile of tadalafil and the latest clinical evi-dence on the management of ED and ED-related comorbidities,aiming to provide a reference for clinical practice.

Objective:To investigate the prognostic factors for esophageal cancer after radical resection and the application value of machine learning prediction model.Methods:The retrospective cohort study was conducted. The clinicopatholigical data of 406 esophageal cancer patients who were admitted to Qilu Hospital of Shandong University from January 2018 to March 2022 were collected. There were 357 males and 49 females, aged (64±8)years. All patients underwent radical resection of esophageal cancer. The 406 patients were randomly divided into a training set of 285 cases and a validation set of 121 cases at a 7∶3 ratio based on a random number table. The training set was used to construct prediction model, and the validation set was used to validate prediction model. Patients were divided into high-risk group and low-risk group based on risk scores. Observation indicators: (1) follow-up of patients and analysis of influencing factors for prognosis; (2) construction and validation of machine learning prediction models. Comparison of measurement data with normal distribution between groups was conducted using the independent sample t test. Comparison of measurement data with skewed distribution between groups was conducted using the Mann-Whitney U test. Comparison of count data between groups was conducted using the chi-square test. Comparison of ordinal data between groups was conducted using the rank sum test. The Kaplan-Meier method was used to calculate survival rate and plot survival curve, and the Log-rank test was used for survival analysis. The Cox proportional hazard regression model was used for univariate and multivariate analyses. Independent influencing factors were included, and data processing, machine learning model construction, and visualization were performed using R packages including random survival forest (RSF), gradient boosting machine (GBM), least absolute shrinkage and selection operator Cox regression (LASSO-Cox), Cox proportional hazards model boosting (CoxBoost), survival support vector machine (survivalsvm), extreme gradient boosting (XGBoost), supervised principal component analysis (SuperPC), and Cox partial least squares regression (plsRcox). Receiver operating characteristic (ROC) curves were drawn, and sensitivity, specificity, and area under the curve (AUC) were calculated. The Delong test was used to assess the differences in AUC among different models in the training set, and the time-dependent ROC was used to compare the predictive performance of different models. Calibration curves were used to evaluate model accuracy, and decision curve analysis (DCA) was used to evaluate overall net benefit. Results:(1) Follow-up of patients and analysis of influencing factors for prognosis. All 406 patients were followed up postoperatively for 28(range, 6-36)months, with 1- and 3-year overall survival rate of 86.5% and 40.9%, respectively. The 285 patients in the training set were followed up postoperatively for 30(range, 6-36)months, with 1- and 3-year overall survival rate of 85.1% and 35.5%, respectively. The 121 patients in the validation set were followed up postoperatively for 25(range, 6-36)months, with 1- and 3-year overall survival rate of 87.0% and 43.2%, respectively. There was no significant difference in postoperative overall survival rate between the training set and the validation set ( χ2=3.20, P>0.05). Results of multivariate analysis showed that left thoracic surgical approach, preopera-tive neutrophil count, vascular invasion, perineural invasion, pathological T2-4 stage, pathological N2-3 stage, and postoperative pneumonia were independent risk factors affecting postoperative survival of 285 patients in the training set ( hazard ratio=1.466, 1.037, 1.482, 1.549, 5.268, 7.727, 22.202, 2.539, 2.686, 1.425, 95% confidence interval as 1.026-2.096, 1.003-1.073, 1.008-2.179, 1.105-2.170, 1.201-23.099, 1.833-32.576, 4.734-104.128, 1.577-4.087, 1.631-4.422, 1.018-1.994, P<0.05). (2) Construction and validation of machine learning prediction models. Independent risk factors affecting postoperative survival were included to construct RSF, GBM, LASSO-Cox, CoxBoost, survivalsvm, XGBoost, SuperPC, and plsRcox machine learning prediction models. Results of Delong test showed that there were significant differences in the AUC of RSF and GBM from the other six models ( P<0.05). Results of time-dependent ROC curve showed that all 8 machine learning predic-tion models had good discriminative ability in the training cohort, among which the RSF machine learning prediction model had the best predictive performance. Results of calibration curve showed that the RSF machine learning prediction model fitted well for predicting postoperative 1-, 2-, and 3-year overall survival in the training cohort, with high consistency with actual results. Results of decision curve analysis showed that within a threshold range of 0-0.80, the RSF machine learning prediction model provided a better overall net benefit. Further analysis showed that in the validation set, the AUC of RSF machine learning prediction model for postoperative 1-, 2-, and 3-year survival prediction were 0.786 (95% confidence interval as 0.609-0.962), 0.774 (95% confidence interval as 0.676-0.873), and 0.750 (95% confidence interval as 0.652-0.848), respectively. Results of calibration curve showed that the RSF machine learning prediction model fitted well for predicting postopera-tive 1-, 2-, and 3-year overall survival in the validation set, with high consistency with actual results. In the training set, the optimal cutoff value of the RSF machine learning prediction model risk score was 11.7. Patients with risk score ≥11.7 were classified as the high-risk group, and those with risk score <11.7 as the low-risk group. The median survival times of the two groups were 18.0 months and >36.0 months, respectively, showing a significant difference between them ( χ2=73.30, P<0.05). In the validation set, the optimal cutoff value of the RSF machine learning prediction model risk score was 11.7. Patients with risk score ≥11.7 were classified as the high-risk group, and those with risk score<11.7 as the low-risk group. The median survival times of the two groups were 17.0 months and>36.0 months for the high-risk and low-risk groups, respectively, showing a significant difference between them ( χ2=35.20, P<0.05). Conclusions:Left thoracic surgical approach, preoperative neutrophil count, vascular invasion, perineural invasion, pathological T2-4 stage, pathological N2-3 stage, and postoperative pneumonia are independent risk factors affecting survival of esophageal cancer patients after radical resection. The RSF machine learning prediction model constructed based on these factors can effectively distinguish the survival prognosis of high-risk and low-risk patients.