Heart sound is one of the common medical signals for diagnosing cardiovascular diseases. This paper studies the binary classification between normal or abnormal heart sounds, and proposes a heart sound classification algorithm based on the joint decision of extreme gradient boosting (XGBoost) and deep neural network, achieving a further improvement in feature extraction and model accuracy. First, the preprocessed heart sound recordings are segmented into four status, and five categories of features are extracted from the signals based on segmentation. The first four categories of features are sieved through recursive feature elimination, which is used as the input of the XGBoost classifier. The last category is the Mel-frequency cepstral coefficient (MFCC), which is used as the input of long short-term memory network (LSTM). Considering the imbalance of the data set, these two classifiers are both improved with weights. Finally, the heterogeneous integrated decision method is adopted to obtain the prediction. The algorithm was applied to the open heart sound database of the PhysioNet Computing in Cardiology(CINC) Challenge in 2016 on the PhysioNet website, to test the sensitivity, specificity, modified accuracy and F score. The results were 93%, 89.4%, 91.2% and 91.3% respectively. Compared with the results of machine learning, convolutional neural networks (CNN) and other methods used by other researchers, the accuracy and sensibility have been obviously improved, which proves that the method in this paper could effectively improve the accuracy of heart sound signal classification, and has great potential in the clinical auxiliary diagnosis application of some cardiovascular diseases.
Algorithms ; Databases, Factual ; Heart Sounds ; Neural Networks, Computer

Objective To report the early mortality and associated risk factors after damage control operation (DCO) in patients with severe abdominal trauma.Methods A total of 146 patients hospitalized from January to March in 2015 were retrospectively analyzed.Early death was defined as death occurring within 72 h of first surgery and before the definitive surgery.Based on the death definition, the subjects were divided into death group and survival group.The two groups were compared for gender, age, injury time, injury classification, vital signs, need for cardiopulmonary resuscitation, Glasgow coma score (GCS), injury severity score (ISS), hemoglobin, platelets, prothrombin time (PT), activated partial thromboplastin time (APTT), blood pH, base excess, operative time, and postoperative acute physiology and chronic health evaluation score Ⅱ (APACHE Ⅱ).Logistic regression analysis was applied to examine the predicators of early postoperative mortality.Results There were 118 patients (80.8％) in survival group and 28 patients (19.2％) in death group.Survival and death groups differed significantly with regard to the percent of elderly (6.8％ vs.21.4％), percent of patients with multiple injury (62.7％ vs.85.7％), body temperature [(36.1 ± 0.4) vs.(35.2 ± 0.8) ℃], percent of patients with greater ISS (31.4％ vs.64.3％), PT [(12.1±1.5) vs.(13.9±1.2)s], bloodpH (7.25±0.04vs.7.08±0.11), base excess [(-8.9±2.8) vs.(-10.6±3.3)mmol/L], postoperative APACHE Ⅱ[(12.8 ± 1.8) vs.(17.5 ± 2.0) points] (P ＜ 0.05).Logistic regression analysis identified the age (OR=1.512, 95％ CI 1.112-4.157,P＜0.05), ISS (OR =1.313,95％ CI 1.134-5.442, P ＜ 0.05), APACHE Ⅱ (OR =1.361,95 ％ CI 1.182-5.222, P ＜ 0.05) as the independent risk factors for early mortality.Conclusion The patients underwent DCO for severe abdominal trauma has a high early mortality, which is closely associated with the age, injury severity and postoperative medical status.

Background:Gastric cancer is a common gastrointestinal malignancy. The detection rate of early gastric cancer is still low in China,and some gastric cancer patients visit the hospital due to acute disease,such as gastric perforation. Aims:To investigate the influence of acute disease on long-term prognosis in gastric cancer patients. Methods:A total of 318 patients with gastric cancer from Jan. 2009 to Jan. 2015 at Shaoxing People’s Hospital were enrolled,and were divided into acute disease group and non-acute disease group. Clinicopathological characteristics were analyzed. Patients were followed up,and survival rate was compared between the two groups. Results:Fifty-three(16. 7% )patients had acute disease,and the remaining 265(83. 3% )patients were without acute disease. Compared with non-acute disease group, percentage of TNM Ⅳ stage was significantly increased(P < 0. 05),and percentage of radical surgery was significantly decreased in acute disease group(P < 0. 05). Kaplan-Meier assay showed that survival rate in acute disease group was significantly decreased when compared with non-acute disease group( P < 0. 05). After adjusting the TNM stage and surgical treatment,no significant difference in survival rate was found between the two groups. Conclusions:Gastric cancer patients with acute disease have lower survival rate. Acute disease may be not an independent prognostic factor, higher TNM stage and lower proportion of radical surgery are the main reasons for poor prognosis in gastric cancer patients with acute disease.

Coronavirus Infections ; epidemiology ; Humans

Objective To analyze the association of the single nucleotide polymorphism (SNP) of Toll-like receptor 7 (TLR7) and Toll-like receptor 9 (TLR9) with chronic hepatitis C virus (HCV)infection.Methods A total of 150 patients with chronic hepatitis C (CHC) admitted to Renmin Hospital of Wuhan University from January 2011 to May 2012 and 168 healthy controls were enrolled in the study.The genotypes of TLR7 IVS2-151 (rs179009) were detected by Sanger sequencing,and the genotypes of TLR9 T-1486C (rs187084) were determined by polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP).SPSS 15.0 was used for statistical analysis,and goodness-of-fit test for HardyWeinberg equilibrium was also performed.Results The frequency of TLR7 IVS2-151G was higher in malepatients with CHC than that in male controls (41.4％ vs.21.6％,x2 =7.250,P =0.007,OR =0.389,95％ CI:0.194-0.781) ; however the female CHC patients had a higher frequency of TLR7 IVS2-151A than the female controls (76.9％ vs.63.1％,x2 =7.202,P =0.007,OR =1.942,95％ CI:1.192-3.164).No significant difference in the distribution of TLR9 T-1486C (rs187084) gene SNP was observed betweenCHC and control groups (P ＞0.05).Conclusion TLR7 IVS2-151 (rs179009) is correlated with HCV infection,which may be involved in the pathogenesis of CHC.

Objective To investigate the formation and special cell biological behavior of osteoclasts.Methods The live-cell imaging technology was adopted to consecutively and dynamically observe the whole process of multinuclear osteoclast formation induced by RANKL and M-CSF from rat peripheral blood monocyte.Meanwhile,the inverted phase contrast microscopy,TRAP staining,and scanning electron microscopy were also applied to identify the osteoclast.Results After 2-week cultivation,a great number of apocytes were found by the inverted phase contrast microscopy,and most apocytes and monocytes had positive reaction after TRAP staining.Moreover,many bone resorption lacunae in which osteoclasts were perhrming bone resorption function could be found in the bone slice under the scanning electron microscope.Live-cell imaging observation showed that the multinuclear osteoclasts were generated through self-fusion of monocytes,fusion of monocytes and apocytes,as well as fusion between apocytes.All fusion processes occurred under the condition of cell adherence.Time-lapse Microcinematography observation showed diverse shapes of osteoclasts and the cell division of multinuclear osteoclasts.Conclusion Rat peripheral blood monocyte can develop into osteoclast under induction of RANKL and M-CSF.Osteoclast can form gigantic apocyte via various types of cell fusion to increase its nucleus number and cell volume,vary its shape,and increase the area of plasma membrane.On the other hand,osteoclast can decrease its cell volume and nucleus number via cell division to adapt the needs of local morphology,biomechanics and bone resorption dynamics.It suggests that this non-mitosis cell division is a special cell biological behavior of osteoclast,which may be the basis of exerting its function and improving bone resorption efficiency.

OBJECTIVETo establish an analytical method on vitexin concentration in plasma to determine the plasma protein binding rate of vitexin.

METHODThe equilibrium dialysis method and HPLC were adopted to determine vitexin concentration in plasma, calculate human plasma protein binding rate of vitexin and compare rat and human plasma protein binding rates of vitexin.

RESULTAt 2-16 mg x L(-1), there was no significant difference in the plasma protein binding rate. But the human plasma protein binding rate of vitexin was higher than its rat plasma protein binding rate, indicating a significant difference in rat and human plasma protein binding rates of vitexin.

CONCLUSIONVitexin has a higher protein binding rate with both rat plasma and human plasma.

Animals ; Apigenin ; metabolism ; Blood Proteins ; metabolism ; Chromatography, High Pressure Liquid ; Dialysis ; Humans ; Protein Binding ; Rats ; Species Specificity

Objective To identify significantly differentially expression of rat by genes chip,try to find out the initiating factors and molecular mechanisms that oxidative stress originate or strengthen the steriod-induced necrosis of femoral head(SINFH).Methods Twenty Wistar rats were dived into experimental group and control group randomly.The rats were injected intraperitoneally whith endotoxin,and then injected intramuscularly with high-dose methylprednisolone or saline in experimental group and control group respectively.The mRNA was extracted from the femoral head of rats inevery group,and the cDNA probes were obtained by inverse transcript,and then carried out microarray detection.The quantitative RT-PCR was used to confirm the results of the microarray.Results Histopathological findings revealed that the experimental group rats had femoral head necrosis,trabecular bone disorders,thinning,bone cell necrosis,and the rate of empty lacunae increased,and in control group no femoral head osteonecrosis was found.Total of 27 differentially expressed genes were found,and of which 4 genes(COX6A2,COX4I2,SOD3,and DUSP1)were significantly different.The expression of these 4 genes were down-regulated.The functions of these four genes involved in inhibition of reactive oxygen species(ROS)generation,accelerated removal of ROS and protection tissue from oxidative damage and so on.Conclusion The expression of oxidative stress-related genes in SINFH of rats exist change.COX6A2,COX4I2,SOD3,and DUSP1 are key genes in process of oxidative stress originate or strengthen the SINFH.

Objective To investigate the effect of nucleocapsid (N) protein of severe acute respiratory syndrome-associated coronavirus (SARS-CoV) on the expression of cyclooxygenase-2 (COX-2). Methods 293T cells were co-transfected with reporter plasmid pCOX-2-Luc containing the luciferase gene under the control of COX-2 promoter and plasmids carrying individual genes of SARS-CoV, and luciferase activity was measured. Expression of COX-2 mRNA and protein was measured by reverse transcription-polymerase chain reaction (RT-PCR) and Western blot. Results N protein of SARS-CoV enhanced COX-2 gene promoter activity, and upregulated COX-2 mRNA expression.COX-2 protein production in 293T cells was N protein concentration-dependent. Conclusion N protein of SARS-CoV could specifically activate COX-2 expression.

Objective Get a primary culture method of osteoblast which is simple to operate and has many pure cells. Method Get osteoblast with improved I collagenase, and test the form of osteoblast, the alkaline phosphatase dyeing and formation of mineralization node. Result The improved rats primary culture method of osteoblast can get most cells in normal shape, alkaline phosphatase secretion and mineralization. Conclusion The improved rats primary culture method of osteoblast is convenient to operate, and can get most cells.