Abstract Survival analysis has been widely used in the field of medical research. The Cox proportional hazard model is commonly used, but its practical application is limited. Machine learning method can compensate for the shortcomings of the Cox proportional hazard model in terms of nonlinear data processing and prediction accuracy. This article reviewed the advance of machine learning methods represented by neural networks, within the field of survival analysis, and highlighted the principles and benefits of three machine learning methods that DeepSurv, Deep-Hit and random survival forest, providing methodological insights for the analysis of complex survival data.

Objective:To analyze the changes rule of serum procalcitonin (PCT) levels in patients with traumatic brain injury in plateau areas, and to evaluate its value in assessing the severity and prognosis of the patients.Methods:A prospective cohort study was conducted. The patients with traumatic brain injury admitted to the critical care medicine departments of Xining Third People's Hospital (at an altitude of 2 260 metres) and Golmud City People's Hospital (at an altitude of 2 780 metres) from May 2018 to September 2022 were enrolled. According to the Glasgow coma scale (GCS) score at admission, the patients were divided into mild injury group (GCS score 13-15), severe injury group (GCS score 9-12), and critical injury group (GCS score 3-8). All patients received active treatment. Chemiluminescence immunoassay was used to measure the serum PCT levels of patients on the 1st, 3rd, 5th, and 7th day of admission. The Kendall tau-b correlation method was used to analyze the correlation between serum PCT levels at different time points and the severity of the disease. The patients were followed up until October 30, 2022. The prognosis of the patients was collected. The baseline data of patients with different prognosis were compared. The Cox regression method was used to analyze the relationship between baseline data, serum PCT levels at different time points and prognosis. Receiver operator characteristic curve (ROC curve) was drawn to analyze the predictive value of serum PCT levels at different time points for death during follow-up.Results:Finally, a total of 120 patients with traumatic brain injury were enrolled, including 52 cases in the mild injury group, 40 cases in the severe injury group, and 28 cases in the critical injury group. The serum PCT levels of patients in the mild injury group showed a continuous downward trend with the prolongation of admission time. The serum PCT levels in the severe injury and critical injury groups reached their peak at 3 days after admission, and were significantly higher than those in the mild injury group (μg/L: 3.53±0.68, 4.47±0.63 vs. 0.40±0.14, both P < 0.05), gradually decreasing thereafter, but still significantly higher than the mild injured group at 7 days. Kendall tau-b correlation analysis showed that there was a significant positive correlation between serum PCT levels on days 1, 3, 5, and 7 of admission and the severity of disease ( r value was 0.801, 0.808, 0.766, 0.528, respectively, all P < 0.01). As of October 30, 2022, 92 out of 120 patients with traumatic brain injury survived and 28 died, with a mortality of 23.33%. Compared with the survival group, the GCS score, serum interleukin-6 (IL-6) levels, white blood cell count (WBC) in peripheral blood, and PCT levels in cerebrospinal fluid at admission in the death group were significantly increased [GCS score: 5.20±0.82 vs. 4.35±0.93, IL-6 (ng/L): 1.63±0.45 vs. 0.95±0.27, blood WBC (×10 9/L): 14.31±2.03 vs. 11.95±1.98, PCT in cerebrospinal fluid (μg/L): 11.30±1.21 vs. 3.02±0.68, all P < 0.01]. The serum PCT levels of patients in the survival group showed a continuous downward trend with prolonged admission time. The serum PCT level in the death group peaked at 3 days after admission and was significantly higher than that in the survival group (μg/L: 4.11±0.62 vs. 0.52±0.13, P < 0.01), gradually decreasing thereafter, but still significantly higher than the survival group at 7 days. Cox regression analysis showed that serum IL-6 levels [hazard ratio ( HR) = 17.347, 95% confidence interval (95% CI) was 5.874-51.232], WBC in peripheral blood ( HR = 1.383, 95% CI was 1.125-1.700), PCT levels in cerebrospinal fluid ( HR = 1.952, 95% CI was 1.535-2.482) at admission and serum PCT levels on admission days 1, 3, 5, and 7 [ HR (95% CI) was 6.776 (1.844-24.906), 1.840 (1.069-3.165), 3.447 (1.284-9.254), and 6.666 (1.214-36.618), respectively] were independent risk factors for death during follow-up in patients with traumatic brain injury (all P < 0.05). ROC curve analysis showed that the AUC of serum PCT levels on days 1, 3, 5, and 7 for predicting death during follow-up in patients with traumatic brain injury was all > 0.8 [AUC (95% CI) was 0.898 (0.821-0.975), 0.800 (0.701-0.899), 0.899 (0.828-0.970), 0.865 (0.773-0.958), respectively], indicating ideal predictive value. The optimal cut-off value for serum PCT level at 3 days of admission was 1.88 μg/L, with the sensitivity of 78.6% and specificity of 88.0% for predicting death during follow-up. Conclusions:Abnormal expression of serum PCT levels in patients with traumatic brain injury on the 3rd day of admission was found. The serum PCT levels greater than 3 μg/L may be related to severe illness. The serum PCT levels greater than 1.88 μg/L can predict the poor prognosis of patients. Dynamic observation of changes in serum PCT levels has good evaluation value for the severity and prognosis of patients with traumatic brain injury in plateau areas.

Osteoporotic proximal humeral fracture (OPHF) is one of the common osteoporotic fractures in the aged, with an incidence only lower than vertebral compression fracture, hip fracture, and distal radius fracture. OPHF, secondary to osteoporosis and characterized by poor bone quality, comminuted fracture pattern, slow healing, and severely impaired shoulder joint function, poses a big challenge to the current clinical diagnosis and treatment. In the field of diagnosis, treatment, and rehabilitation of OPHF, traditional Chinese and Western medicine have accumulated rich experience and evidence from evidence-based medicine and achieved favorable outcomes. However, there is still a lack of guidance from a relevant consensus as to how to integrate the advantages of the two medical systems and achieve the integrated diagnosis and treatment. To promote the diagnosis and treatment of OPHF with integrated traditional Chinese and Western medicine, relevant experts from Orthopedic Expert Committee of Geriatric Branch of Chinese Association of Gerontology and Geriatrics, Youth Osteoporosis Group of Orthopedic Branch of Chinese Medical Association, Osteoporosis Group of Orthopedic Surgeon Branch of Chinese Medical Doctor Association, and Osteoporosis Committee of Shanghai Association of Integrated Traditional Chinese and Western Medicine have been organized to formulate Expert consensus on the diagnosis and treatment of osteoporotic proximal humeral fracture with integrated traditional Chinese and Western medicine ( version 2024) by searching related literatures and based on the evidences from evidence-based medicine. This consensus consists of 13 recommendations about the diagnosis, treatment and rehabilitation of OPHF with integrated traditional Chinese medicine and Western medicine, aimed at standardizing, systematizing, and personalizing the diagnosis and treatment of OPHF with integrated traditional Chinse and Western medicine to improve the patients ′ function.

The incidence of intracranial aneurysms is higher in population, and the death and disability rates after rupture are extremely high. Early detection of intracranial aneurysms, prediction of rupture risk, operation program, and evaluation of treatment outcomes are all the clinical concerns. With the continuously deepening integration of artificial intelligence technology and medicine, its application in the diagnosis and treatment of intracranial aneurysms is becoming increasingly widespread. This article reviews the current application status of artificial intelligence technology in the diagnosis and treatment of intracranial aneurysms.

Objective:To investigate the efficacy and safety of Tubridge flow diverter (TFD) in the treatment of unruptured intracranial aneurysms (UIAs).Methods:The clinical data of patients with UIAs treated with TFD in the Department of Neurosurgery, Linyi People's Hospital from July 2021 to July 2023 were retrospectively analyzed. The modified Rankin Scale (mRS) was used to evaluate functional outcomes during postoperative follow-up. Digital subtraction angiography (DSA) follow-up was performed at least 3 months after procedure, and the occlusion degree of the aneurysm was assessed with the O'Kelly-Marotta (OKM) grading system.Results:A total of 52 patients (71 aneurysms) were included and 54 TFDs were successfully implanted. Forty-one patients (78.8%) were treated with TFD alone, while 11 (21.2%) were treated with TFD combined with coil embolization. One patient experienced in-stent thrombosis during the procedure, and the symptoms disappeared after medical treatment. Postoperative hemorrhage and ischemic complications occurred in each 2 cases, respectively. No patient died during the follow-up period. DSA follow-up was performed on a total of 45 aneurysms in 34 patients. The followed-up time was 8.3±3.1 months (range, 3-15 months), including 2 OKM grade A (4.4%), 11 grade B (24.4%), 1 grade C (2.2%), and 31 grade D (68.9%). At the last follow-up, 51 patients had good outcomes (mRS score 0-2), and 1 patient had poor outcome (mRS score 4).Conclusion:TFD for UIAs has a high complete occlusion rate and a low complication rate.

Gastric cancer with hemorrhage and cerebral infarction is a serious complica-tion with poor prognosis in clinic. Although the incidence rate is extremely low, the fatality and disability rates are very high. In addition, the opposition in treatment between the two complica-tions increases the difficulty of clinical diagnosis and treatment. The authors report the diagnosis and treatment of a gastric cancer patient with hemorrhage and new cerebral infarction, in order to to provide reference for related treatments.

Objective:To compare the safety and efficacy of ticagrelor and clopidogrel in dual antiplatelet therapy for stent-assisted embolization of unruptured intracranial aneurysms.Methods:Patients with unruptured intracranial aneurysms received stent-assisted embolization in the Department of Neurosurgery, Linyi People's Hospital from January 2021 to June 2022 were retrospectively included. According to the preprocedural dual antiplatelet therapy scheme, they were divided into aspirin+clopidogrel group (clopidogrel group) and aspirin+ticagrelor group (ticagrelor group). The incidence of ischemic and bleeding events was compared between the clopidogrel group and the ticagrelor group at 3 months after procedure. Multivariate logistic regression model was used to analyze independent risk factors for postprocedural ischemic and bleeding events. Results:A total of 195 patients were included. Their age was 58.15±10.11 years and 75 were males (38.5%). There was no statistically significant difference in the incidence of bleeding events (12.8% vs. 5.9%) and ischemic events (14.9% vs. 18.8%) at 3 months after procedure between the ticagrelor group ( n=94) and the clopidogrel group ( n=101). Multivariate logistic regression analysis showed that smoking (odds ratio [ OR] 6.085; 95% confidence interval [ CI] 1.589-13.012; P=0.019], hypertension ( OR 4.547, 95% CI 1.589-13.012; P=0.005), aneurysm at the branch vessel ( OR 3.089, 95% CI 1.122-8.504; P=0.029), and the use of flow diverter ( OR 3.111, 95% CI 1.062-9.110; P=0.038) were the independent risk factors for postprocedural ischemic events. Triglycerides might be an independent risk factor for postprocedural bleeding events ( OR 1.435, 95% CI 0.989-2.082; P=0.057), but did not reach statistical significance. Conclusions:In dual antiplatelet therapy for stent-assisted embolization of unruptured intracranial aneurysms, ticagrelor and clopidogrel have the same safety and efficacy.

Objective:To assess freshmen's perception of medical university learning environment, and to provide evidence and suggestions for the university to improve its learning environment in a targeted way.Methods:From March to April 2020, JHLES (Johns Hopkins University Learning Environment Scale) questionnaires were distributed to 260 freshmen from a military medical university. The influence factors of students' scores were analyzed by one-way analysis of variance (ANOVA) and general linear model. SAS 9.4 software was used for statistical analysis.Results:A total of 244 valid questionnaires were collected, with a response rate of 93.85%(244/260). The average score of inclusion and safety dimension was the lowest (2.21±1.08), while the average score of community of peers' dimension was the highest (4.41±0.54). There were significant differences among gender, whether one-child, parents' education level, overall learning environment evaluation and learning interest (all P<0.05). Gender ( P<0.05), learning interest ( P<0.05) and maternal education ( P<0.05) were main influencing factors of students' learning environment perception. Conclusion:Medical universities should further encourage good relationship among students, and enhance students' interest in learning through various measures, especially for female students and students who has relative poorer maternal education.

Objective:To explore the application value of ventricular intracranial pressure monitoring (V-ICPM) in the treatment of unilateral temporal lobe cerebral contusion.Methods:A retrospective cohort study was conducted to analyze the clinical data of 295 patients with unilateral temporal lobe cerebral contusion admitted to 904th Hospital of PLA Joint Support Force from January 2014 to August 2021, including 172 males and 123 females; aged 14-78 years [(46.3±14.7)years]. V-ICPM was used in 136 patients (V-ICPM group), who received surgical or non-surgical treatment according to the monitoring, while not in 159 patients (non-V-ICPM group), who received routine surgery or non-surgical treatment. The two groups were compared in terms of the rates of intracranial hematoma clearance by craniotomy, decompressive craniectomy (DC) and dehydration and osmotic therapy during hospitalization, use time of 20% mass fraction of mannitol and 30 g/L hypertonic salt, displacement rate of brain midline structure of head CT≥10 mm after discharge, rate of intracranial infection, hydrocephalus and epilepsy, and Glasgow Outcome Scale (GOS) at 6 months after discharge.Results:All patients were followed up for 6-12 months [(8.9±2.1)months]. During hospitalization, the rate of intracranial hematoma clearance by craniotomy and the rate of DC in V-ICPM group were 35.3% (48/136) and 8.1% (11/136), lower than 47.2% (75/159) and 22.0% (35/159) in non-V-ICPM group ( P<0.05 or 0.01). There was no significant difference between the two groups in the rate of dehydration and osmotic therapy or the use time of mannitol (all P>0.05). The use time of hypertonic salt in V-ICPM group was (7.2±2.5)days, more than (4.1±1.8)days in non-V-ICPM group ( P<0.05). After discharge, the displacement rate of brain midline structure of head CT in V-ICPM group was 29.4% (40/136), lower than 42.8% (68/159) in non-V-ICPM group ( P<0.05). There was no significant difference between the two groups in the rate of intracranial infection, hydrocephalus and epilepsy (all P>0.05). Six months after discharge, the good rate of GOS in V-ICPM group was 91.2% (124/136), significantly better than 81.8% (130/159) in non-V-ICPM group ( P<0.05). Conclusion:For unilateral temporal lobe cerebral contusion, V-ICPM is associated with reduced rate of craniotomy exploration and DC, decreased incidence of complications and improved prognosis of the patients in spite of longer use time of hypertonic salt.

The condition of patients with severe traumatic brain injury (sTBI) complicated by corona virus 2019 disease (COVID-19) is complex. sTBI can significantly increase the probability of COVID-19 developing into severe or critical stage, while COVID-19 can also increase the surgical risk of sTBI and the severity of postoperative lung lesions. There are many contradictions in the treatment process, which brings difficulties to the clinical treatment of such patients. Up to now, there are few clinical studies and therapeutic norms relevant to sTBI complicated by COVID-19. In order to standardize the clinical treatment of such patients, Critical Care Medicine Branch of China International Exchange and Promotive Association for Medical and Healthcare and Editorial Board of Chinese Journal of Trauma organized relevant experts to formulate the Chinese expert consensus on clinical treatment of adult patients with severe traumatic brain injury complicated by corona virus infection 2019 ( version 2023) based on the joint prevention and control mechanism scheme of the State Council and domestic and foreign literatures on sTBI and COVID-19 in the past 3 years of the international epidemic. Fifteen recommendations focused on emergency treatment, emergency surgery and comprehensive management were put forward to provide a guidance for the diagnosis and treatment of sTBI complicated by COVID-19.