Objective To predict the risk of in-hospital death in patients with chronic heart failure(CHF)complicated by lung infections using interpretable machine learning.Methods The clinical data of 1415 patients diagnosed with CHF complicated by lung infections were obtained from the MIMIC-IV database.According to the pathogen type,the patients were categorized into bacterial pneumonia and non-bacterial pneumonia groups,and their risks of in-hospital death were compared using Kaplan-Meier survival curves.Univariate analysis and LASSO regression were used to select the features for constructing LR,AdaBoost,XGBoost,and LightGBM models,and their performance was compared in terms of accuracy,precision,F1 value,and AUC.External validation of the models was performed using the data from eICU-CRD database.SHAP algorithm was applied for interpretive analysis of XGBoost model.Results Among the 4 constructed models,the XGBoost model showed the highest accuracy and F1 value for predicting the risk of in-hospital death in CHF patients with lung infections in the training set.In the external test set,the XGBoost model had an AUC of 0.691(95%CI:0.654-0.720)in bacterial pneumonia group and an AUC of 0.725(95%CI:0.577-0.782)in non-bacterial pneumonia group,and showed better predictive ability and stability than the other models.Conclusion The overall performance of the XGBoost model is superior to the other 3 models for predicting the risk of in-hospital death in CHF patients with lung infections.The SHAP algorithm provides a clear interpretation of the model to facilitate decision-making in clinical settings.

Objective To predict the risk of in-hospital death in patients with chronic heart failure(CHF)complicated by lung infections using interpretable machine learning.Methods The clinical data of 1415 patients diagnosed with CHF complicated by lung infections were obtained from the MIMIC-IV database.According to the pathogen type,the patients were categorized into bacterial pneumonia and non-bacterial pneumonia groups,and their risks of in-hospital death were compared using Kaplan-Meier survival curves.Univariate analysis and LASSO regression were used to select the features for constructing LR,AdaBoost,XGBoost,and LightGBM models,and their performance was compared in terms of accuracy,precision,F1 value,and AUC.External validation of the models was performed using the data from eICU-CRD database.SHAP algorithm was applied for interpretive analysis of XGBoost model.Results Among the 4 constructed models,the XGBoost model showed the highest accuracy and F1 value for predicting the risk of in-hospital death in CHF patients with lung infections in the training set.In the external test set,the XGBoost model had an AUC of 0.691(95%CI:0.654-0.720)in bacterial pneumonia group and an AUC of 0.725(95%CI:0.577-0.782)in non-bacterial pneumonia group,and showed better predictive ability and stability than the other models.Conclusion The overall performance of the XGBoost model is superior to the other 3 models for predicting the risk of in-hospital death in CHF patients with lung infections.The SHAP algorithm provides a clear interpretation of the model to facilitate decision-making in clinical settings.

Objective:To establish a method for rapid detection of OXA and par C resistance genes of Acinetobacter baumannii(Ab)by double nucleic acid colloidal gold strip and to develop kit.Methods:DNA of Ab was extracted by heating and boiling method.OXA and par C genes sequences of Ab were selected as target gene fragments based on NCBI.Primers were designed and labeled with 6-FAM,digoxin and biotin,respectively.Drug resistance gene detection reagents were developed,and nucleic acid gold test strips were used for rapid and visual detection.Molecular cloning and sequencing techniques were used to clone positive control samples and evaluate specificity,sensitivity and stability of kit.Results:DNA concentration and purity of Ab extracted by boiling method were good.Homology between cloned and sequenced plasmid DNA and gene sequence in GenBank database was 100%,respectively.Speci-ficity of kit was good,with only Ab showing positive results and other bacterial genera showing negative results;DNA concentration of Ab in double nucleic acid colloidal gold test strip decreased to 10-3 ng/μl,a red line still appeared,whose sensitivity was 10 times higher consistent with minimum detection limit of electrophoresis 10-2 ng/μl;test kits were tested at 3rd,6th and 9th months,and showed good stability.Conclusion:Double resistance detection kit established in this study can simultaneously detect OXA and par C resis-tance of Ab,who has advantages of high sensitivity,strong specificity,rapid and simple,and provides a new method for detection of carbapenem and quinolone antibiotic resistance of Ab.

Objective:To explore the clinicopathological features and prognosis of renal mucinous tubular spindle cell carcinoma (MTSCC).Methods:The clinical data of 16 patients with MTSCC admitted to the First Affiliated Hospital of Zhengzhou University from July 2013 to July 2022 were retrospectively analyzed. There were six male cases and ten female cases. The mean age was (56.4±11.4) years old. Among them, 10 cases were asymptomatic, two complained of hematuria, three complained of lumbar pain, and one complained of lower limb pain. Twelve cases underwent preoperative enhanced CT examination, 6 cases of ultrasound examination, 3 cases of MRI examination, and 1 case of bone scan. Imaging manifestations showed that the masses were round or round-like with clear borders. Two cases combined with hemorrhage and three cases combined with calcification. Five cases showed exophytic growth, 10 cases partially exophytic, and 1 case completely endophytic. The maximum diameter of the tumor was (65.7±27.4) mm. The tumors were located in the left kidney in 11 cases and in the right kidney in 5 cases. The tumors were mildly delayed-enhancing under enhanced CT, long/short T1 signal mixed with long/short T2 signal under MRI, and diffusion-limited high signal under DWI. The tumors were hypoechoic masses without obvious blood flow signals under ultrasound. Twelve cases were diagnosed as renal occupying neoplasms, 2 cases were suggested as lack of blood supply renal tumor, and one was considered renal tumor rupture and bleeding. In one case, a bone scan suggested metastasis to the thoracic spine and pelvis. The metastatic renal tumor was diagnosed, and a renal puncture was performed to clarify the pathology. Eleven patients underwent laparoscopic radical nephrectomy, and 4 patients underwent partial nephrectomy. One case was metastasized without surgery and treated with apatinib mesylate and zoledronic acid.Results:The postoperative pathological specimens showed grayish, grayish-yellow, or grayish-red masses with a soft or medium texture. No perinephric, ureteral, or adrenal invasion was seen in all tumors. Microscopically, the tumor cells were round and ovoid. The tumor cells were arranged in tubular and striated shapes, and mucus pools were locally visible. No sarcomatous component was seen in all tumors. There were 9 patients with pT 1N 0M 0, 6 patients with pT 2N 0M 0, and 1 patient with pT 1N 0M 1. After operation, 2 patients with pT 2N 0M 0, who underwent laparoscopic radical nephrectomy were treated with pazopanib and sunitinib, respectively. All patients were followed up for a median of 50.7(25.8, 75.0)months, 15 patients were free of recurrent metastases, and 1 patient with pre-puncture metastasis died due to tumor progression of multiple pulmonary and bone metastases, with a survival of 16.9 months. Conclusions:Renal MTSCC is rare, mostly found on physical examination, with female patients predominantly, and imaging shows a lack of blood supply tumor. Surgery is the primary treatment method. Partial nephrectomy or radical nephrectomy could be chosen according to the tumor stage, kidney function, and patient's underlying condition, and patients have a good prognosis.

Objective To discuss the clinical effect of perioperative bundle of care for patients with advanced hepatocellular carcinoma (HCC) who are receiving CT-guided percutaneous minimally-invasive argon-helium knife cryoablation.Methods A total of 30 HCC patients,who underwent percutaneous argonhelium knife cryoablation,were enrolled in this study.Perioperative measures based on the concept of bundle of care were implemented,which included training of nursing team members,perfect preoperative psychological nursing,dietary guidance,preoperative routine preparations,effective intraoperative guidance and close observation of the patient's condition,postoperative activity guidance,prevention and observation of complications,etc.Results All 30 advanced HCC patients could actively cooperate with physicians for the performance of percutaneous argon helium cryoablation.After the treatment,two patients developed nausea and vomiting and one patient developed chills and fever,which were improved after symptomatic treatment.All the 30 patients recovered well during the perioperative period and were discharged from hospital smoothly.Conclusion Perioperative bundle of care can help patients restore the surgical damage as soon as possible,reduce the pain and improve the quality of life.

With the increasing incidence of osteoporosis, osteoporotic vertebral compression fracture (OVCF) has become a common disease in the elderly. Relevant researches have indicated that percutaneous vertebroplasty (PVP) has become the main treatment for painful osteoporotic vertebral compression fractures. In aspect of perioperative nursing care for PVP, both domestic and foreign scholars have published a large number of research reports, indicating that perioperative nursing intervention can alleviate the pain, reduce the incidence of complications, improve the quality of life of patients, and promote rehabilitation. This article aims to make a comprehensive review about the perioperative nursing care for patients with osteoporotic vertebral compression fractures who are receiving PVP so as to provide reference for clinical nursing staff in the implementation of nursing intervention. (J Intervent Radiol, 2018, 27:290-293)

Objective: To study the segment of liver according to the large amount of three-dimensional(3D) reconstructive images of normal human livers and the vascular system, and to recognize the basic functional liver unit based on the anatomic features of the intrahepatic portal veins. Methods: The enhanced CT primitive DICOM files of 1 260 normal human livers from different age groups who treated from October 2013 to February 2017 provided by 16 hospitals were analyzed using the computer-aided surgery system.The 3D liver and liver vascular system were reconstructed, and the digital liver 3D model was established.The vascular morphology, anatomical features, and anatomical distributions of intrahepatic portal veins were statistically analyzed. Results: The digital liver model obtained from the 3D reconstruction of CAS displayed clear intrahepatic portal vein vessels of level four.Perform a digital liver segments study based on the analysis of level four vascular distribution areas.As the less anatomical variation of left hepatic portal vein, the liver was classified into four types of liver segmentation mainly based on right hepatic portal vein.Type A was similar to Couinaud or Cho′s segmentation, containing 8 segments(537 cases, 42.62%). Type B contained 9 segments as there are three ramifications of right-anterior portal vein(464 cases, 36.82%). The main difference for Type C was the variation of right-posterior portal vein which was sector shape(102 cases, 8.10%). Type D contained the cases with special portal vein variations, which needs three-dimensional simulation to design individualized liver resection plan(157 cases, 12.46%). These results showed that there was no significant difference in liver segmental typing between genders(χ²=2.179, P=0.536) and did not reveal any significant difference in liver segmental typing among the different age groups(χ²=0.357, P=0.949). Conclusions The 3D digital liver model can demonstrate the true 3D anatomical structures, and its spatial vascular variations.The observation of anatomic features, distribution areas of intrahepatic portal veins and individualized liver segmentation achieved via digital medical 3D visualization technology is of great value for understand the complexity of liver anatomy and to guide the precise hepatectomy.

Objective To discuss the perioperative nursing measures for patients with osteoporotic vertebral compression fractures who are receiving percutaneous vertebroplasty (PVP) treatment by using high viscosity bone cement.Methods A total of 30 patients with osteoporotic vertebral compression fractures were included in this study.All patients were treated with PVP by using high viscosity bone cement.Preoperative routine nursing,psychological intervention,dietary intervention,postoperative guidance of rehabilitation exercise of limbs,close observation of bone cement leakage were strictly implemented,and the corresponding nursing measures were promptly taken when needed.Results Through strict implement of the nursing intervention all 30 patients could actively cooperate with PVP treatment,and after PVP the pain was significantly relieved in all patients.Conclusion Adequate preoperative preparation,proper postoperative guidance,careful observation and effective nursing can help the patients resume their daily life activities as soon as possible,relieve the pain,and improve the quality of life as well.(J Intervent Radiol,2017,26:274-276)

Objective To explore optimal initial the best screening time for newborns with different delivery methods using AABR.Methods A total of 550 newborns who were born from August 1, 2016 to October 31, 2016 at our hospital participated in the study.AABR was used to accomplish the initial hearing screening.The newborns were divided into 2 groups according to the delivery methods.There were each 100 neonates born in vaginal during <24 h, 24~48 h and 48~72 h after birth, respectively.The numbers of neonates delivered by cesarean section during the 3 separate periods were 50, 100 and 100, respectively.The newborns who failed the preliminary hearing screening proceeded to the re-screening and diagnostic procedures.Results There were 300 newborns were born in vaginal, and the pass rate in 24~48 h after birth group was significantly higher than that in 24 h group (93.00% vs 83.00%,x2=4.735,P=0.03<0.05), but it was not significantly different from that of 48~72 h group (95.00% vs 93.00%,x2=0.355,P=0.56>0.05).There were 250 newborns in cesarean section, the pass rate of 24~48 h after birth group was significantly higher than that in 24 h group (83.00% vs 68.00%,x2=4.437, P=0.04<0.05), and significantly lower than that of 48~72 h group (94.00% vs 83.00%,x2=5.944, P=0.02<0.05).Conclusion Taking into account of hospitalization time, the screening time for the vaginal delivery newborn hearing screening can be advanced to 24~48 h after birth with the application of AABR, but not for the cesarean section group.

BACKGROUNDBased on the excellent medical care and management system for Chinese veterans, as well as the detailed medical documentation available, we aim to construct a Chinese Veteran Clinical Research (CVCR) platform on non-communicable diseases (NCDs) and carry out studies of the primary disabling NCDs.

METHODSThe Geriatric Neurology Department of Chinese People's Liberation Army General Hospital and veterans' hospitals serve as the leading and participating units in the platform construction. The fundamental constituents of the platform are veteran communities. Stratified typical cluster sampling is adopted to recruit veteran communities. A cross-sectional study of mental, neurological, and substance use (MNS) disorders are performed in two stages using screening scale such as the Mini-Mental State Examination and Montreal cognitive assessment, followed by systematic neuropsychological assessments to make clinical diagnoses, evaluated disease awareness and care situation.

RESULTSA total of 9 676 among 277 veteran communities from 18 cities are recruited into this platform, yielding a response rate of 83.86%. 8 812 subjects complete the MNS subproject screening and total response rate is 91.70%. The average participant age is (82.01±4.61) years, 69.47% of veterans are 80 years or older. Most participants are male (94.01%), 83.36% of subjects have at least a junior high school degree. The overall health status of veterans is good and stable. The most common NCD are cardiovascular disorders (86.44%), urinary and genital diseases (73.14%), eye and ear problems (66.25%), endocrine (56.56%) and neuro-psychiatric disturbances (50.78%).

CONCLUSIONWe first construct a veterans' comprehensive clinical research platform for the study of NCDs that is primarily composed of highly educated Chinese males of advanced age and utilize this platform to complete a cross-sectional national investigation of MNS disorders among veterans. The good and stable health condition of the veterans could facilitate the long-term follow-up studies of NCDs and provide prospective data to the prevention and management of NCDs.

Aged ; Aged, 80 and over ; Asian Continental Ancestry Group ; Cross-Sectional Studies ; Disease ; Female ; Health Status ; Humans ; Male ; Veterans ; statistics & numerical data