OBJECTIVES: To evaluate the performance of a multi-constraint representation learning classification model for identifying ovarian cancer with missing laboratory indicators. METHODS: Tabular data with missing laboratory indicators were collected from 393 patients with ovarian cancer and 1951 control patients. The missing ovarian cancer laboratory indicator features were projected to the latent space to obtain a classification model using the representational learning classification model based on discriminative learning and mutual information coupled with feature projection significance score consistency and missing location estimation. The proposed constraint term was ablated experimentally to assess the feasibility and validity of the constraint term by accuracy, area under the ROC curve (AUC), sensitivity, and specificity. Cross-validation methods and accuracy, AUC, sensitivity and specificity were also used to evaluate the discriminative performance of this classification model in comparison with other interpolation methods for processing of the missing data. RESULTS: The results of the ablation experiments showed good compatibility among the constraints, and each constraint had good robustness. The cross-validation experiment showed that for identification of ovarian cancer with missing laboratory indicators, the AUC, accuracy, sensitivity and specificity of the proposed multi-constraints representation-based learning classification model was 0.915, 0.888, 0.774, and 0.910, respectively, and its AUC and sensitivity were superior to those of other interpolation methods. CONCLUSIONS The proposed model has excellent discriminatory ability with better performance than other missing data interpolation methods for identification of ovarian cancer with missing laboratory indicators.
Female ; Humans ; Ovarian Neoplasms/diagnosis* ; Machine Learning ; ROC Curve

Accurate classification of the acetabular injuries and appropriate treatment plan are great challenges for orthopedic surgeons because of the irregular anatomical structure of the acetabulum and aggregation of important vessels and nerves around it. Letournel-Judet classification system has been widely applied to classify acetabular fractures. However, there are several limitations, including incomplete inclusion of fracture types, difficulty in understanding and insufficient guidance for surgical treatment, etc. Serious complications such as traumatic arthritis are common due to wrong classification and diagnosis and improper selection of surgical strategy, which brings a heavy burden to the society and families. Three-column classification, based on anatomic characteristics, has advantages of containing more fracture types and being easy to understand, etc. To solve the problems existing in the diagnosis and treatment process based on Letournel-Judet classification, achieve accurate diagnosis and treatment of patients with acetabular fractures, and obtain satisfactory prognosis, the Orthopedic Trauma Emergency Center of Third Hospital of Hebei Medical University and the Trauma Orthopedic Branch of the Chinese Orthopedic Association organized experts from relevant fields to formulate the Expert consensus on the accurate diagnosis and treatment of acetabular fractures based on three-column classification ( version 2023) in terms of principles of evidence-based medicine. Based on the three-column classification, 15 recommendations were proposed, covering the diagnosis, treatment, complication prevention and management, etc, so as to provide reference for accurate diagnosis and treatment of acetabular fractures.

Objective:To investigate the effects of naturally occurring mild hypothermia and artificial mild hypothermia on the expression of high mobility group box 1(HMGB1) in lung tissues of septic mice.Methods:One hundred and twenty BALB/C mice (SPF level) were randomly numbered.Twelve mice with integer multiples of 10 were used as the normal control (NC) group, and the remaining 108 mice were chosen as the septic group.The septic mouse model was established by intra abdominal injection of lipopolysaccharide (LPS) 10 mg/kg.The NC group was given the same dose of normal saline.Anal temperature of the septic mice were measured 1 hour after the model was established successfully, and then were divided into naturally occurring mild hypothermia group and non-mild hypothermia group according to T≤36℃ and T>36℃.In the naturally occurring mild hypothermia group, the mice with T<34℃ were eliminated, and the remaining septic mice were randomly divided into the naturally occurring mild hypothermia(NOMH) observation group and the keep normothermia (KN) group.NOMH group was not given preheating intervention, while KN group was placed in an incubator to maintain the anal temperature between 36.0℃ and 37.5℃.Septic mice in the non-mild hypothermia group were randomly divided into the nonhypothermia (NH) observation group and the artificial mild hypothermia (ATMH) group.The NH group was not treated with hypothermia, while the ATMH group was treated with physical hypothermia, so that the anal temperature of the mice were maintained at 34℃-36℃.Four mice in each group were randomly selected at 6 and 12 hours after modeling, and the concentrations of tumor necrosis factor-α(TNF-α), interleukin-6(IL-6) and HMGB1 in serum were detected by enzyme-linked immunosorbent assay(ELISA). At 12 hours, the survival rate of each group of mice was observed.Then 4 mice of each group were sacrificed and lung tissues were taken.The pathological changes of lung tissues were observed by hematoxylin-eosin (HE) staining, and the expression of HMGB1 in lung tissues was observed by immunohistochemical staining.Real time fluorescence quantitative PCR and Western blot were used to detect the relative expression of HMGB1 at mRNA and protein levels.Results:(1)Twelve hours after modeling, the survival number of NOMH group, ATMH group, KN group and NH group were 36(40), 6(11), 27(40), 4(11), respectively, and there were differences between the four groups (χ 2=32.286, P=0.002). Compared with the other three groups of septic mice, the survival rate was highest in the NOMH group (compared with ATMH group: χ 2=5.222, P=0.022; compared with the KN group: χ 2=6.050, P=0.013; and the NH group: χ 2=11.672, P=0.001), but the differences between the other two groups were not statistically significant (all P>0.05). (2)Compared with the NC group, the concentrations of serum TNF-α, IL-6 and HMGB1 of septic mice in each group were significantly increased at 6 h and 12 h (all P<0.05). Compared with NOMH group, the concentrations of TNF-α, IL-6 and HMGB1 in ATMH group, KN group and NH group were significantly increased at 6 h and 12 h(all P<0.05), and the concentrations of TNF-α, IL-6 and HMGB1 in NH group were the highest at all time points (all P<0.05). The concentrations of TNF-α at 12 h decreased compared with 6 h (all P<0.05), while the concentrations of IL-6 and HMGB1 at 12 h increased compared with 6 h (all P<0.05). (3)HE staining showed that the lung tissue damage were minimal in NOMH group, followed by ATMH group.(4)Immunohistochemical staining showed that the expression of HMGB1 protein was in order of NOMH group, ATMH group, KN group and NH group; (5)The relative expressions of HMGB1 protein in lung tissues of septic mice in NOMH group, ATMH group, KN group, and NH group was 0.280±0.013, 0.320±0.016, 0.340±0.018, and 0.380±0.014, respectively, and the relative expression level of HMGB1 mRNA was 4.86±0.22, 6.02±0.18, 6.26±0.20, and 7.98±0.28, respectively, compared with NC group (HMGB1 protein content was 0.240±0.013, and the relative expression level of HMGB1 mRNA was 2.21±0.12) significantly increased (all P<0.05). Cmpared with NOMH group, the relative expression levels of HMGB1 protein and HMGB1 mRNA in the lung tissues of the ATMH group, KN group and NH group were significantly increased(all P<0.05), with the highest expression level in the NH group(all P<0.05). Conclusion:Mild hypothermia may reduce lung tissue damage by down-regulating the expression of HMGB1 in lung tissues of septic mice, and the improvement of spontaneous mild hypothermia was more significant.

OBJECTIVE:To standardize hospital purchasing management for medicinal material. METHODS:The separation of the four powers,including plan,purchase,supply,and payment,was practiced on the basis of centralized purchasing medicinal material in our hospital. SWOT method was used to analyze the strengths,weaknesses,opportunities and threats of the model;im-provement measures were formulated;finally the effects of the model were evaluated. RESULTS:This model played an active role in the professional advantage of purchasing management,made up for the lack of management system and manpower shortage dis-advantage;adapted to policies to seize the opportunities for improvement,reduced the threat of standard deletion and purchasing cost,and promoted reasonable consumption of medicinal materials. And then the purchasing cost of our hospital was reduced after system procedure improvement,fine management of medicinal material and professional level improvement,medicine ratio fell from 42% in 2010 to 37% in 2014;logistics support became stable and standard,and the level of medicinal material supply was improved. CONCLUSIONS:The purchasing model for medicinal material under separation of the four powers provides the refer-ence for the correct course of the standardized,normalized and modernized hospital purchasing management.