ABSTRACT
Ultrasound examination has the advantages of non-radiation, non-invasive, low cost and high efficiency, and is the most commonly used method of liver imaging examination. In recent years, the application of computer vision technology to the intelligent analysis of ultrasound images has become a research hotspot in the field of intelligent healthcare. Through large-scale data training, the intelligent analysis model of ultrasound omics based on machine learning algorithm can assist clinical diagnosis and therapy, and improve the efficiency and accuracy of diagnosis. Based on the literature, the authors summarize the application proprect of computer vision technology assisted ultrasonography in the evaluation of diffuse liver lesions, focal liver lesions, microvascular invasion of liver cancer, postoperative recurrence of liver cancer, and postoperative therapy response to trans-catheter arterial chemoembolization.
ABSTRACT
In recent years, the artificial intelligence machine learning and deep learning technology have made leap progress. Using clinical decision support system for auxiliary diagnosis and treatment is the inevitable developing trend of wisdom medical. Clinicians tend to ignore the interpretability of models while pursuing its high accuracy, which leads to the lack of trust of users and hamper the application of clinical decision support system. From the perspective of explainable artificial intelligence, the authors make some preliminary exploration on the construction of clinical decision support system in the field of liver disease. While pursuing high accuracy of the model, the data governance techniques, intrinsic interpretability models, post-hoc visualization of complex models, design of human-computer interactions, providing knowledge map based on clinical guidelines and data sources are used to endow the system with interpretability.
ABSTRACT
ObjectiveTo investigate the application of Mengchao Liver Disease-Brain System version 2.0 in clinical diagnosis and treatment. MethodsThis study was conducted among 160 patients who were admitted to the internal medicine and surgical departments from June 9 to 21， 2021， and their data were automatically captured by the intelligent information system of Southeast Big Data Institute of Hepatobiliary Health， Mengchao Hepatobiliary Hospital of Fujian Medical University. The completeness and accuracy of Mengchao Liver Disease-Brain System version 2.0 were evaluated based on the intelligent diagnostic tools such as auxiliary diagnosis of chronic hepatitis B， interpretation of liver fibrosis， staging model of chronic hepatitis B， auxiliary diagnosis of liver cirrhosis， auxiliary staining of liver cirrhosis， auxiliary diagnosis of primary liver cancer， BCLC stage of primary liver cancer， Chinese staging of primary liver cancer， Child-Pugh score， and APRI score. ResultsAll auxiliary diagnostic tools had a complete rate of 94.17% in terms of the extraction of correct key dimensions within the test period. The artificial intelligence report had a structured accuracy of 97.55% in capturing data and an accuracy rate of 91.61% in text processing. ConclusionMengchao Liver Disease-Brain System version 2.0 provides an innovative mode for the construction of big data platform in medical specialties and has a high accuracy as an auxiliary diagnostic tool in clinical diagnosis and treatment.
ABSTRACT
Deep learning is a process in which machine learning obtains new knowledge and skills by simulating the learning behavior of human brain through massive data training and analysis. With the development of medical technology, a large amount of data has been accumulated in the medical field, and the research on data may help to understand the relationships and rules within data and predict the onset and prognosis of human diseases. Deep learning can find the hidden information in data and has been increasingly used in the medical field. Primary liver cancer is a malignant tumor with high incidence and mortality rates, poor prognosis, and a high recurrence rate, and early diagnosis, timely treatment, and prediction of recurrence have always been the research hotspots in recent years. This article reviews the advances in the application of deep learning in the diagnosis and recurrence of liver cancer from the aspects of risk prediction, postoperative recurrence, and survival risk prediction.
ABSTRACT
Bioinformatics is an interdisciplinary science that combines the tools of mathematics, computer science, and biology to clarify and explore the biological implications of large amounts of biological data. With the continuous development of genome sequencing technology, a large number of biological data has been generated, and mining of the biological significance contained in big data has become one of the main tasks that need to be solved urgently. This article summarizes the risk prediction models for hepatocellular carcinoma (HCC) based on feature genes, so as to provide new perspectives for early identification, prognosis, and treatment optimization of HCC.
ABSTRACT
Primary liver cancer is one of the most common and fatal malignant tumors, and surgical treatment is the most important radical treatment method, but there is still a high postoperative recurrence rate and poor prognosis. In recent years, emerging techniques represented by artificial intelligence have achieved rapid innovation and are gradually integrated into the whole process of the diagnosis and treatment of primary liver cancer. Promoting the implementation of artificial intelligence in the surgical treatment of primary liver cancer is of great significance to the high-quality development of precision liver surgery. At present, researchers have extensively explored the application of artificial intelligence in treatment decision-making, preoperative evaluation, surgical implementation, postoperative management, and adjuvant therapy for primary liver cancer. This article reviews the advances in the application of artificial intelligence in the surgical treatment of primary liver cancer, so as to accelerate the application of artificial intelligence in clinical diagnosis and treatment, improve clinical service ability, and ultimately improve patients' prognosis.
ABSTRACT
Surgical resection is currently the main method for the treatment of primary liver cancer. The appearance of new techniques, such as three-dimensional visualization, 3D printing, virtual reality, indocyanine green molecular fluorescence imaging, and hepatectomy with intraoperative navigation, has provided new methods for the preoperative diagnosis, surgical planning, and intraoperative navigation of primary liver cancer. Among these techniques, three-dimensional visualization shows incomparable advantages in the diagnosis of primary liver cancer, the selection of treatment regimen, preoperative planning, intraoperative navigation, and liver transplantation. This article summarizes the recent advances in the application of three-dimensional visualization in surgical operation for primary liver cancer.
ABSTRACT
Objective:To clarify the characteristics of renal cortical microcirculation and its relationship with the expression of plasma endothelial microparticle (EMP) in septic rats, and to evaluate the effect of Xuebijing injection as an adjuvant therapy of antibiotics on septic AKI.Methods:The 8-10 weeks old specific pathogen free (SPF) male Sprague-Dawley (SD) rats were divided into sham operation group (Sham group), positive drug control group and Xuebijing group by the random number table method, with 10 rats in each group. The cecal ligation and puncture (CLP) with large ligation (ligated 75% of the cecum) was used to prepare a rat high-grade sepsis model; in the Sham group, the cecum was stretched without ligation or puncture. Due to the high mortality of CLP with large ligation, Xuebijing injection (4 mL/kg, 12 hours per time) and imipenem/cilastatin injection (90 mg/kg, 6 hours per time) were administered to the rats in the Xuebijing group via the tail vein immediately after the model was produced. Normal saline and imipenem/cilastatin were administered to the rats by the same methods in the positive drug control group. The rats in the Sham group were treated with the same volume of normal saline as any of the other two groups at the same frequency. At 48 hours after model reproduction, the mean arterial pressure (MAP) and blood lactic acid (Lac) of the rats were measured. The renal cortical microcirculation was monitored by using side stream dark-field imaging. Renal hypoxia signals were assessed by pimonidazole chloride immunohistochemistry. Plasma EMP levels were determined by using flow cytometry, and then the correlation between EMP and microcirculation parameters of renal cortex was analyzed. At the same time, the serum creatinine (SCr) was measured, and the renal injury score (Paller score) was used to evaluate the severity of renal tissue pathological damage.Results:Compared with the Sham group, perfused vessel density (PVD), microvascular flow index (MFI) and MAP in the positive drug control group and the Xuebijing group decreased significantly, the positive expression of hypoxia probe (pimonidazole) increased, Lac, EMP, Paller score and SCr increased significantly. However, compared with the positive drug control group, the renal cortical microcirculation in the Xuebijing group was improved significantly, PVD and MFI were increased significantly [PVD (mm/mm 2): 16.20±1.20 vs. 9.77±1.12, MFI: 2.46±0.05 vs. 1.85±0.15, both P < 0.05], Lac was reduced significantly (mmol/L: 4.81±1.23 vs. 6.08±1.09, P < 0.05), MAP increased slightly [mmHg (1 mmHg = 0.133 kPa): 84.00±2.00 vs. 80.00±2.00, P > 0.05], suggested that Xuebijing injection improved renal microcirculation perfusion in septic rats, and this effect did not depend on the change of MAP. The positive expression of pemonidazole in renal cortex of the Xuebijing group was significantly lower than that of the positive drug control group [(35.89±1.13)% vs. (44.93±1.37) %, P < 0.05], suggested that Xuebijing injection alleviated renal hypoxia. The plasma EMP levels of rats in the Xuebijing group were significantly lower than those in the positive drug control group (×10 6/L: 3.49±0.17 vs. 5.78±0.22, P < 0.05), and the EMP levels were significantly negatively correlated with PVD and MFI ( r values were -0.94 and -0.95, respectively, both P < 0.05), indicated that the increase of plasma EMP was highly correlated with renal microcirculation disorder, and Xuebijing injection inhibited the increase of plasma EMP levels. The Paller score in the Xuebijing group was significantly lower than that in the positive drug control group (46.90±3.84 vs. 62.70±3.05, P < 0.05), and the level of SCr was also significantly lower than that in the positive drug control group (μmol/L: 121.1±12.4 vs. 192.7±23.9, P < 0.05), which suggested that Xuebijing injection relieved kidney injury and improved renal function in septic rats. Conclusion:As an adjuvant therapy of antibiotics, Xuebijing injection could inhibit the expression of plasma EMP in rats with sepsis, improve renal cortex microcirculation, and reduce kidney injury.
ABSTRACT
Objective:To investigate the long-term effect of open reduction and internal fixation and one stage rotator cuff repair in treatment of proximal humeral fractures combined with rotator cuff tear.Methods:A retrospective case series study was performed on clinical data of 40 patients with proximal humeral fractures combined with rotator cuff tear admitted to Ruijin Hospital Affiliated to Shanghai Jiaotong University School of Medicine from January 2012 to December 2015. There were 24 males and 16 females, aged 52-93 years [(72.5±10.2)years]. According to Neer classification, there were 7 patients with two-part proximal humerus fractures, 27 with three-part proximal humerus fractures and 6 with four-part proximal humerus fractures. All patients received open reduction and internal fixation using locking plates and one stage rotator cuff repair. Functional shoulder exercises were carried out after operation. Fracture healing and wound healing were observed postoperatively. Shoulder active motion, visual analogue score (VAS) and University of California at Los Angeles (UCLA) shoulder rating scale were assessed and compared before operation and at the last follow-up.Results:All patients were followed up for 4.5-8.5 years [(6.2±1.2)years]. All wounds and fractures healed without infection or nonunion. At the last follow-up, the active forward elevation of affected shoulder was 100°-150° [(121.8±15.8)°], the external rotation with the shoulder in neutral position was 30°-80° [(47.0±15.9)°], the external rotation with the shoulder abduction was 60°-80° [(73. 5±6.2)°], the internal rotation with the shoulder abduction was 60°-80° [(70.5±7.1)°], showing significant differences compared to preoperative 30°-60° [(44.8±11.1)°], 0°-30° [(12.0±10.4)°], 0°-30° [(13.8±7.7)°], 30°-60° [(47.0±8.5)°], respectively (all P<0.01). The VAS was 1 (0, 1)points at the last follow-up, and was 6(5, 6)points before operation ( P<0.01). The UCLA shoulder rating scale was 28-35 points [(31.0±2.3)points] at the last follow-up, compared with preoperative 8-11 points [(10.3±0.8)points] ( P<0.01), and rated as excellent in 10 patients, good in 23 and poor in 7, with the excellent-good rate of 83%. Conclusion:For patients with proximal humeral fractures combined with rotator cuff tear, open reduction and internal fixation and one stage rotator cuff repair can significantly restore the shoulder motion, relieve the pain and recovery shoulder function, and the long-term therapeutic effect is satisfactory.
ABSTRACT
Objective@#To investigate the clinical value of lymph node dissection (LND) in the radical resection of intrahepatic cholangiocarcinoma (ICC).@*Methods@#The propensity score matching and retrospective cohort study was conducted. The clinicopathological data of 448 patients with ICC who were admitted to 12 medical centers from December 2011 to December 2017 were collected, including 279 in the Eastern Hepatobiliary Surgery Hospital of Navy Medical University, 32 in the Mengchao Hepatobiliary Hospital of Fujian Medical University, 21 in the First Hospital Affiliated to Army Medical University, 20 in the Cancer Hospital Chinese Academy of Medical Science and Peking Union Medical College, 19 in the West China Hospital of Sichuan University, 18 in the Second Hospital Affiliated to Zhejiang University School of Medicine, 18 in the Renji Hospital Affiliated to Shanghai Jiaotong University School of Medicine, 16 in the Beijing Friendship Hospital Affiliated to Capital Medical University, 10 in the Xuanwu Hospital Affiliated to Capital Medical University, 7 in the Tongji Hospital Affiliated to Tongji Medical College of Huazhong University of Science and Technology, 5 in the Beijing Tiantan Hospital Affiliated to Capital Medical University, and 3 in the Affiliated Hospital of North Sichuan Medical College. There were 281 males and 167 females, aged from 22 to 80 years, with a median age of 57 years. Of the 448 patients, 143 with routinely intraoperative LND were divided into LND group and 305 without routinely intraoperative LND were divided into control group, respectively. Observation indicators: (1) the propensity score matching conditions and comparison of general data between the two groups after matching; (2) intraoperative and postoperative situations; (3) follow-up; (4) survival analysis. Patients were followed up by outpatient examination, telephone interview and email to detect survival of patients and tumor recurrence up to October 31, 2018 or death. Follow-up was conducted once every 3 months within postoperative 1-2 years, once every 6 months within postoperative 2-5 years, and once a year after 5 years. The propensity score matching was realized using the nearest neighbor method with 1∶1 ratio. Measurement data with normal distribution were represented as Mean±SD, and comparison between groups was analyzed using the t test. Measurement data with skewed distribution were represented as M (range), and comparison between groups was analyzed using the Mann-Whitney U test. Count data were described as absolute numbers, and comparison between groups was analyzed using the chi-square test or Fisher exact probability. Kaplan-Meier method was used to calculate survival rates and draw survival curve, and Log-rank test was used to perform survival analysis. Univariate analysis and multivaraiate analysis were conducted using the Log-rank test and COX regression model, respectively.@*Results@#(1) The propensity score matching conditions and comparison of general data between the two groups after matching: 286 of 448 patients had successful matching, including 143 in each group. The number of males and females, cases with hepatitis, cases with grade A and B of Child-Pugh classification, cases with distance from margin to tumor <1 cm and ≥1 cm, cases with highly, moderately, and poorly differentiated tumor, cases with anatomical hepatectomy and nonanatomical hepatectomy before propensity score matching were 77, 66, 33, 96, 47, 70, 73, 105, 38, 79, 64 in the LND group, and 204, 101, 121, 165, 140, 207, 98, 251, 54, 124, 181 in the control group, respectively, showing significant differences in the above indices between the two groups (χ2=7.079, 11.885, 7.098, 14.763, 5.184, 8.362, P<0.05). After propensity score matching, the above indices were 77, 66, 33, 96, 47, 70, 73, 105, 38, 79, 64 in the LND group, and 79, 64, 29, 88, 55, 71, 72, 112, 31, 74, 69 in the control group, respectively, showing no significant difference in the above indices between the two groups (χ2=0.056, 0.329, 2.158, 0.014, 0.936, 1.400, P>0.05). (2) Intraoperative and postoperative situations: the operation time, volume of intraoperative blood loss, cases with intraoperative blood transfusion, cases with postoperative complications, and duration of postoperative hospital stay were 265 minutes (range, 160-371 minutes), 300 mL (range, 200-500 mL), 37, 46, 12 days (range, 9-17 days) for the LND group, and 59 minutes (range, 46-250 minutes), 200 mL (range, 100-400 mL), 24, 25, 9 days (range, 7-11 days) for the control group, respectively. There was no significant difference in the volume of intraoperative blood loss or cases with intraoperative blood transfusion between the two groups (Z=1.700, χ2=3.520, P>0.05). There were significant differences in the operation time, cases with postoperative complications, and duration of postoperative hospital stay (Z=6.520, χ2=8.260, Z=4.270, P<0.05). (3) Follow-up: 143 patients in the LND group was followed up for 18-26 months. The 1-, 3-, 5-year survival rates were 66.9%, 32.8%, 25.0%, and the median survival time was 22.0 months (range, 18.0-26.0 months). In the control group, 143 patients were followed up for 11-32 months. The 1-, 3-, 5-year survival rates were 71.7%, 38.0%, 31.0%, and the median survival time was 22.0 months (range, 11.0-32.0 months). There was no significant difference in the overall survival between the two groups (χ2=0.466, P>0.05). (4) Survival analysis: results of univariate analysis showed that distance from margin to tumor, tumor diameter, the number of tumors, microvascular tumor thrombus, and lymph node metastasis by pathological examination were related factors for prognosis of patients undergoing radical resection of ICC (χ2=10.662, 9.477, 16.944, 6.520, 11.633, P<0.05). Results of multivariate analysis showed that distance from margin to tumor<1 cm, tumor diameter>5 cm, multiple tumors, microvascular tumor thrombus, positive lymph node metastasis by pathological examination were independent risk factors for prognosis of patients undergoing radical resection of ICC (hazard ratio=0.600, 1.571, 1.601, 1.750, 1.723, 95% confidence interval: 0.430-0.837, 1.106-2.232, 1.115-2.299, 1.083-2.829, 1.207-2.460, P<0.05).@*Conclusion@#Lymph node metastasis is an independent risk factor affecting the prognosis of ICC patients, and it is necessary to perform LND in radical resection for patients with ICC and lymph node metastasis.
ABSTRACT
With its high efficiency for site-specific genome editing and easy manipulation, the clustered regularly interspaced short palindromic repeats (CRISPR)/ CRISPR associated protein 9 (CAS9) system has become the most widely used gene editing technology in biomedical research. In addition, significant progress has been made for the clinical development of CRISPR/CAS9 based gene therapies of human diseases, several of which are entering clinical trials. Here we report that CAS9 protein can function as a genome mutator independent of any exogenous guide RNA (gRNA) in human cells, promoting genomic DNA double-stranded break (DSB) damage and genomic instability. CAS9 interacts with the KU86 subunit of the DNA-dependent protein kinase (DNA-PK) complex and disrupts the interaction between KU86 and its kinase subunit, leading to defective DNA-PK-dependent repair of DNA DSB damage via non-homologous end-joining (NHEJ) pathway. XCAS9 is a CAS9 variant with potentially higher fidelity and broader compatibility, and dCAS9 is a CAS9 variant without nuclease activity. We show that XCAS9 and dCAS9 also interact with KU86 and disrupt DNA DSB repair. Considering the critical roles of DNA-PK in maintaining genomic stability and the pleiotropic impact of DNA DSB damage responses on cellular proliferation and survival, our findings caution the interpretation of data involving CRISPR/CAS9-based gene editing and raise serious safety concerns of CRISPR/CAS9 system in clinical application.
ABSTRACT
With its high efficiency for site-specific genome editing and easy manipulation, the clustered regularly interspaced short palindromic repeats (CRISPR)/ CRISPR associated protein 9 (CAS9) system has become the most widely used gene editing technology in biomedical research. In addition, significant progress has been made for the clinical development of CRISPR/CAS9 based gene therapies of human diseases, several of which are entering clinical trials. Here we report that CAS9 protein can function as a genome mutator independent of any exogenous guide RNA (gRNA) in human cells, promoting genomic DNA double-stranded break (DSB) damage and genomic instability. CAS9 interacts with the KU86 subunit of the DNA-dependent protein kinase (DNA-PK) complex and disrupts the interaction between KU86 and its kinase subunit, leading to defective DNA-PK-dependent repair of DNA DSB damage via non-homologous end-joining (NHEJ) pathway. XCAS9 is a CAS9 variant with potentially higher fidelity and broader compatibility, and dCAS9 is a CAS9 variant without nuclease activity. We show that XCAS9 and dCAS9 also interact with KU86 and disrupt DNA DSB repair. Considering the critical roles of DNA-PK in maintaining genomic stability and the pleiotropic impact of DNA DSB damage responses on cellular proliferation and survival, our findings caution the interpretation of data involving CRISPR/CAS9-based gene editing and raise serious safety concerns of CRISPR/CAS9 system in clinical application.
ABSTRACT
Objective:To systematically review the clinical effectiveness of combined postoperative adjuvant transcatheter arterial chemoembolization (TACE) with portal vein chemotherapy (PVC) versus TACE alone in patients with hepatocellular carcinoma (HCC) and portal vein tumor thrombus (PVTT).Methods:Databases including PubMed, Embase, Cochrane Library, Medline, Web of Science, CNKI, China Biology Medicine, Wan Fang and VIP were searched from Jan 1st 2000 to Jun 30th 2019 for eligible studies on clinical effectiveness of combined postoperative adjuvant TACE with PVC versus TACE alone in patients with HCC and PVTT. The primary endpoints were overall survival (OS) and disease-free survival (DFS). The secondary endpoint was adverse events. These endpoints were evaluated by odds ratio ( OR) and 95% confidence interval ( CI) using Review Manager 5.3 software. Results:Nine studies with 642 patients were enrolled in this meta-analysis. There were 323 patients in the TACE group and 319 patients in the TACE plus PVC group. The pooled OR for the 1-, 2-, and 3-year OS were all significantly better in the TACE plus PVC group than the TACE group (1-year OS, OR=2.20, 95% CI: 1.53-3.17; 2-year OS, OR=2.44, 95% CI: 1.69-3.53; 3-year OS, OR=2.30, 95% CI: 1.52-3.46) (all P<0.05). Similarly, significantly better results were observed in the pooled OR for the 1-, 2-, and 3-year DFS (1-year DFS, OR=2.56, 95% CI: 1.70-3.86; 2-year DFS, OR=2.27, 95% CI: 1.19-4.32; 3-year DFS, OR=3.03, 95% CI: 1.55-5.92) (all P<0.05). There were no significant differences in the incidences of adverse events between the two groups (all P>0.05). Conclusion:Postoperative adjuvant TACE combined with PVC for patients with HCC and PVTT was safe and effective, and was significantly better than TACE alone in long-term prognosis. Large-scale, multi-center, prospective studies are needed to support the conclusion.
ABSTRACT
Objective:To investigate the application value of machine learning algorithms for preoperative prediction of microvascular invasion (MVI) in hepatocellular carcinoma (HCC).Methods:The retrospective and descriptive study was conducted. The clinicopathological data of 277 patients with HCC who were admitted to Mengchao Hepatobiliary Hospital of Fujian Medical University between May 2015 and December 2018 were collected. There were 235 males and 42 females, aged (56±10)years, with a range from 33 to 80 years. Patients underwent preoperative magnetic resonance imaging examination. According to the random numbers showed in the computer, all the 277 HCC patients were divided into training dataset consisting of 193 and validation dataset consisting of 84, with a ratio of 7∶3. Machine learning algorithms, including logistic regression nomogram, support vector machine (SVM), random forest (RF), artificial neutral network (ANN) and light gradient boosting machine (LightGBM), were used to develop models for preoperative prediction of MVI. Observation indicators: (1) analysis of clinicopathological data of patients in the training dataset and validation dataset; (2) analysis of risk factors for tumor MVI of the training dataset; (3) construction of machine learning algorithm prediction models and comparison of their accuracy of preoperative tumor MVI prediction. Measurement data with normal distribution were represented as Mean± SD, and comparison between groups was analyzed using the paired t test. Count data were described as absolute numbers, and comparison between groups was analyzed using the chi-square test. Univariate and multivariate analyses were performed using the Logistic regression model. Results:(1) Analysis of clinicopathological data of patients in the training dataset and validation dataset: there were 157 males and 36 females in the training dataset, 78 males and 6 females in the validation dataset, showing a significant difference in the sex between the training dataset and validation dataset ( χ2=6.028, P<0.05). (2) Analysis of risk factors for tumor MVI of the training dataset: of the 193 patients, 108 had positive MVI, and 85 had negative MVI. Results of univariate analysis showed that age, the number of tumors, tumor diameter, satellite lesions, tumor margin, alpha fetaprotein (AFP), alkaline phosphatase (ALP), fibrinogen were related factors for tumor MVI [ odds ratio ( OR)=0.971, 2.449, 1.368, 4.050, 2.956, 4.083, 2.532, 1.996, 95% confidence interval ( CI): 0.943-1.000, 1.169-5.130, 1.180-1.585, 1.316-12.465, 1.310-6.670, 2.214-7.532, 1.016-6.311, 1.323-3.012, P<0.05]. Results of multivariate analysis showed that AFP>20 μg/L, multiple tumors, larger tumor diameter, unsmooth tumor margin were independent risk factors for tumor MVI ( OR=3.680, 3.100, 1.438, 3.628, 95% CI: 1.842-7.351, 1.334-7.203, 1.201-1.721, 1.438-9.150, P<0.05). Larger age was associated with lower risk of preoperative tumor MVI ( OR=0.958, 95% CI: 0.923-0.994, P<0.05). (3) Construction of machine learning algorithm prediction models and comparison of their accuracy of preoperative tumor MVI prediction: ①machine learning algorithm prediction models involving logistic regression nomogram, SVM, RF, ANN and LightGBM were constructed based on results of multivariate analysis including age, AFP, the number of tumors, tumor diameter, tumor margin, and consistency analysis of the logistic regression nomogram prediction model showed a good stability. For the training dataset and validation dataset, the area under curve (AUC) of logistic regression nomogram model, SVM model, RF model, ANN model, LightGBM model was 0.812, 0.794, 0.807, 0.814, 0.810 and 0.784, 0.793, 0.783, 0.803, 0.815, respectively, showing no significant difference between SVM model and logistic regression nomogram model, between RF model and logistic regression nomogram model, between ANN model and logistic regression nomogram model, between LightGBM model and logistic regression nomogram model [(95% CI: 0.731-0.849, 0.744-0.860, 0.752-0.867, 0.747-0.862, Z=0.995, 0.245, 0.130, 0.102, P>0.05) and (95% CI: 0.690-0.873, 0.679-0.865, 0.702-0.882, 0.715-0.891, Z=0.325, 0.026, 0.744, 0.803, P>0.05)]. ② Clinicopathological factors were selected using RF, LightGBM machine learning algorithm to construct corresponding prediction models. According to importance scale of factors to prediction models, factors with importance scale>0.01 were selected to construct RF model, including age, tumor diameter, AFP, white blood cell, platelet, total bilirubin, aspartate transaminase, γ-glutamyl transpeptidase, ALP, and fibrinogen. Factors with importance scale>5.0 were selected to construct LightGBM model, including age, tumor diameter, AFP, white blood cell, ALP, and fibrinogen. Due to lack of factor selection ability, factors based on results of univariate analysis were secected to construct SVM model and ANN model, including age, the number of tumors, tumor diameter, satellite lesions, tumor margin, AFP, ALP, and fibrinogen. For the training dataset and validation dataset, the AUC of SVM model, RF model, ANN model, LightGBM model was 0.803, 0.838, 0.793, 0.847 and 0.810, 0.802, 0.802, 0.836, respectively, showing no significant difference between SVM model and logistic regression nomogram model, between RF model and logistic regression nomogram model, between ANN model and logistic regression nomogram model, between LightGBM model and logistic regression nomogram model [(95% CI: 0.740-0.857, 0.779-0.887, 0.729-0.848, 0.789-0.895, Z=0.421, 0.119, 0.689, 1.517, P>0.05) and (95% CI: 0.710-0.888, 0.700-0.881, 0.701-0.881, 0.740-0.908, Z=0.856, 0.458, 0.532, 1.306, P>0.05)]. Conclusion:Machine learning algorithms can predict MVI of HCC preoperatively, but its application value needs to be further verified by large sample data from multi centers.
ABSTRACT
Objective To investigate the clinical value of lymph node dissection (LND) in the radical resection of intrahepatic cholangiocarcinoma (ICC).Methods The propensity score matching and retrospective cohort study was conducted.The clinicopathological data of 448 patients with ICC who were admitted to 12 medical centers from December 2011 to December 2017 were collected,including 279 in the Eastern Hepatobiliary Surgery Hospital of Navy Medical University,32 in the Mengchao Hepatobiliary Hospital of Fujian Medical University,21 in the First Hospital Affiliated to Army Medical University,20 in the Cancer Hospital Chinese Academy of Medical Science and Peking Union Medical College,19 in the West China Hospital of Sichuan University,18 in the Second Hospital Affiliated to Zhejiang University School of Medicine,18 in the Renji Hospital Affiliated to Shanghai Jiaotong University School of Medicine,16 in the Beijing Friendship Hospital Affiliated to Capital Medical University,10 in the Xuanwu Hospital Affiliated to Capital Medical University,7 in the Tongji Hospital Affiliated to Tongji Medical College of Huazhong University of Science and Technology,5 in the Beijing Tiantan Hospital Affiliated to Capital Medical University,and 3 in the Affiliated Hospital of North Sichuan Medical College.There were 281 males and 167 females,aged from 22 to 80 years,with a median age of 57 years.Of the 448 patients,143 with routinely intraoperative LND were divided into LND group and 305 without routinely intraoperative LND were divided into control group,respectively.Observation indicators:(1) the propensity score matching conditions and comparison of general data between the two groups after matching;(2) intraoperative and postoperative situations;(3) follow-up;(4) survival analysis.Patients were followed up by outpatient examinafion,telephone interview and email to detect survival of patients and tumor recurrence up to October 31,2018 or death.Follow-up was conducted once every 3 months within postoperative 1-2 years,once every 6 months within postoperative 2-5 years,and once a year after 5 years.The propensity score matching was realized using the nearest neighbor method with 1∶1 ratio.Measurement data with normal distribution were represented as Mean±SD,and comparison between groups was analyzed using the t test.Measurement data with skewed distribution were represented as M (range),and comparison between groups was analyzed using the Mann-Whitney U test.Count data were described as absolute numbers,and comparison between groups was analyzed using the chi-square test or Fisher exact probability.Kaplan-Meier method was used to calculate survival rates and draw survival curve,and Log-rank test was used to perform survival analysis.Univariate analysis and multivaraiate analysis were conducted using the Log-rank test and COX regression model,respectively.Results (1) The propensity score matching conditions and comparison of general data between the two groups after matching:286 of 448 patients had successful matching,including 143 in each group.The number of males and females,cases with hepatitis,cases with grade A and B of Child-Pugh classification,cases with distance from margin to tumor ＜ 1 cm and ≥ 1 cm,cases with highly,moderately,and poorly differentiated tumor,cases with anatomical hepatectomy and nonanatomical hepatectomy before propensity score matching were 77,66,33,96,47,70,73,105,38,79,64 in the LND group,and 204,101,121,165,140,207,98,251,54,124,181 in the control group,respectively,showing significant differences in the above indices between the two groups (x2 =7.079,11.885,7.098,14.763,5.184,8.362,P＜0.05).After propensity score matching,the above indices were 77,66,33,96,47,70,73,105,38,79,64 in the LND group,and 79,64,29,88,55,71,72,112,31,74,69 in the control group,respectively,showing no significant difference in the above indices between the two groups (x2 =0.056,0.329,2.158,0.014,0.936,1.400,P＞0.05).(2) Intraoperative and postoperative situations:the operation time,volume of intraoperative blood loss,cases with intraoperative blood transfusion,cases with postoperative complications,and duration of postoperative hospital stay were 265 minutes (range,160-371 minutes),300 mL (range,200-500 mL),37,46,12 days (range,9-17 days) for the LND group,and 59 minutes (range,46-250 minutes),200 mL (range,100-400 mL),24,25,9 days (range,7-11 days) for the control group,respectively.There was no significant difference in the volume of intraoperative blood loss or cases with intraoperative blood transfusion between the two groups (Z =1.700,x2 =3.520,P ＞ 0.05).There were significant differences in the operation time,cases with postoperative complications,and duration of postoperative hospital stay (Z=6.520,x2=8.260,Z=4.270,P＜0.05).(3) Follow-up:143 patients in the LND group was followed up for 18-26 months.The 1-,3-,5-year survival rates were 66.9％,32.8％,25.0％,and the median survival time was 22.0 months (range,18.0-26.0 months).In the control group,143 patients were followed up for 11-32 months.The 1-,3-,5-year survival rates were 71.7％,38.0％,31.0％,and the median survival time was 22.0 months (range,11.0-32.0 months).There was no significant difference in the overall survival between the two groups (x2 =0.466,P＞0.05).(4) Survival analysis:results of univariate analysis showed that distance from margin to tumor,tumor diameter,the number of tumors,microvascular tumor thrombus,and lymph node metastasis by pathological examination were related factors for prognosis of patients undergoing radical resection of ICC (x2 =10.662,9.477,16.944,6.520,11.633,P＜0.05).Results of multivariate analysis showed that distance from margin to tumor＜ 1 cm,tumor diameter＞5 cm,multiple tumors,microvascular tumor thrombus,positive lymph node metastasis by pathological examination were independent risk factors for prognosis of patients undergoing radical resection of ICC (hazard ratio =0.600,1.571,1.601,1.750,1.723,95％ confidence interval:0.430-0.837,1.106-2.232,1.115-2.299,1.083-2.829,1.207-2.460,P＜0.05).Conclusion Lymph node metastasis is an independent risk factor affecting the prognosis of ICC patients,and it is necessary to perform LND in radical resection for patients with ICC and lymph node metastasis.
ABSTRACT
Objective To explore the effects of flipped classroom on the experiment teaching of emergency and critical care nursing. Methods The nursing undergraduates in grade 2014 of Shandong University of Traditional Chinese Medicine were randomly divided into the experimental group and the control group. The traditional teaching model was used in the control group, while the experimental group adopted the flipped classroom teaching model. Students in the experimental group watch micro videos for preview before class , carry out review , discussion , answering , group exercise , guidance , displaying , communication and teacher commenting in class, and undergo summarizing and evaluation after class. The theoretical and operational test scores of the two groups were compared, and the evaluation data of the experimental group students on the effect of flipped classroom teaching model were collected. Results The theoretical scores and experimental scores of students in the experimental group were higher than those in the control group respectively [(82.51 ±5.03) vs. (77.36 ±5.48) and (94.35 ±3.16) vs. (88.02 ±4.27), respectively], and the differences were statistically significant (P=0.011, 0.000). A total of 92.1％ of the students in the experimental group said they liked the teaching model, and 90.3％wanted to continue using the flipped classroom teaching mode . Conclusion The flipped classroom teaching model is highly regarded by the students , and the application of it is helpful to improve students' comprehensive performance.
ABSTRACT
Objective To evaluate different inflammation markers for predicting the risk of acute kidney injury (AKI) in sepsis patients. Methods A retrospective observational study was conducted. The adult patients with sepsis for more than 24 hours admitted to intensive care unit (ICU) of Beijing Friendship Hospital, Capital Medical University from March 1st, 2010 to November 1st, 2017 were enrolled. Inflammatory markers such as white blood cell count (WBC), erythrocyte sedimentation rata (ESR), neutrophil and lymphocyte ratio (NLR), C-reactive protein (CRP), procalcitonin (PCT), etc. were collected at ICU admission. The patients were divided into different groups according to the 24-hour AKI, and subgroup analysis was conducted according to the criteria of the Global Renal Disease Prognostic Organization (KDIGO)-AKI staging. The receiver operating characteristic curve (ROC) was plotted with statistically significant inflammatory markers to assess the predictive value of AKI for patients with systemic infection. Results A total of 753 patients with sepsis were enrolled. 405 AKI patients were diagnosed within 24 hours, with a prevalence of 53.8%. There were 118 cases (15.7%) in AKI stage 1, 48 cases (6.4%) in stage 2 and 239 cases (31.7%) in stage 3. PCT in AKI group was significantly higher than that in non-AKI group [μg/L: 4.98 (1.51, 32.75) vs. 3.00 (0.37, 11.40), P < 0.01]. PCT of AKI stage 2 group was significantly higher than that of AKI stage 1 and 3 groups [μg/L: 27.86 (4.80, 37.26) vs. 3.00 (0.98, 16.10), 4.98 (1.51, 42.55), both P < 0.01]. Although ESR in AKI group was lower than that of non-AKI group (mm/1 h: 45.25±37.42 vs. 52.28±34.89, P < 0.01), there was no significant difference among the subgroups. CRP in AKI group was slightly higher than the non-AKI group [mg/L: 96.00 (42.20, 160.00) vs. 73.60 (21.01, 157.50)], but the difference was not statistically significant (P > 0.05). There was no significant difference in WBC or NLR between AKI group and non-AKI group. It was shown by ROC curve analysis that the area under ROC curve (AUC) of PCT in predicting the occurrence of septic AKI was 0.619, with the 95% confidence interval (95%CI) of 0.545-0.689 (P < 0.01). When the cut-off value of PCT > 0.4 μg/L, the sensitivity was 94.2%, the specificity was 26.5%, the accuracy was 64.2%, the positive predictive value was 61.6%, and the negative predictive value was 78.6%. Conclusion PCT could be a marker to predict AKI with sepsis patients.
ABSTRACT
Objective To investigate the application value of three-dimensional (3D) reconstruction virtual surgery planning in the surgical treatment of hilar cholangiocarcinoma.Methods The retrospective crosssectional study was conducted.The clinical data of 36 patients with hilar cholangiocarcinoma who were admitted to the First Affiliated Hospital of Fujian Medical University between January 2014 and September 2017 was collected.Before operation,images of 3D virtual surgery planning were respectively reconstructed and determined using IQQA-Liver imaging analysis system,and then precise resection of tumor was performed based on results of virtual surgical planning and intraoperative conditions.Observation indicators:(1) virtual surgical planning;(2) surgical and postoperative situations;(3) follow-up and survival.Follow-up using outpatient examination and telephone interview was performed to detect postoperative survival and tumor recurrence or metastasis up to November 2017.Measurement data with normal distribution were represented as (x)±s.Comparison between indicator of preoperative virtual surgical planning and surgical indicators was analyzed by the t test.The postoperative survival time was calculated by the Kaplan-Meier method.Results (1) Virtual surgical planning:36 patients accomplished 3D visualization reconstruction and virtual surgical planning.Three D visualization reconstruction clearly showed adjacent relationship between tumor size and surrounding vessels or bile duct space.Type Ⅱ,Ⅲ a,Ⅲb and Ⅳ of Bismuth-Corlette Classification were detected in 2,13,14 and 7 patients by 3D visualization system,respectively.The tumor volume,whole liver volume,predicted liver resection volume and remnant liver volume were respectively (76± 26) mL,(1 319± 306) mL,(588± 128) mL and (731± 269) mL.(2) Surgical and postoperative situations:of 36 patients,16,12,5 and 3 patients underwent left hemigepatectomy,right hemigepatectomy,extended left hemigepatectomy and extended right hemigepatectomy,respectively,and all of them were combined with caudate lobectomy of liver.Combined resection and reconstruction of hepatic artery,combined wedge resection and repair of the portal vein and combined end-to-end anastomosis after resection of the portal vein were detected in 1,2 and 1 patients,respectively.Operation time and volume of intraoperative blood loss of 36 patients were respectively (368± 134)minutes and (474±288)mL.Thirty-six patients with postoperative complications were cured by conservative treatment,including 3 with pulmonary infection,3 with intra-abdominal infection and 2 with intra-abdominal lymphatic fistula.Duration of hospital stay of 36 patients was (19±7) days.Type Ⅱ,Ⅲ a,Ⅲ b and Ⅳ of postoperative Bismuth-Corlette Classification were detected in 2,11,13 and 10 patients,respectively.Accuracy of tumor classification through 3D visualization reconstruction was 91.7％ (33/36).Actual liver resection volume of 36 patients was (551± 141)mL,and was not significantly different from predicted liver resection volume (t =1.148,P＞0.05).(3) Follow-up and survival:31 of 36 patients were followed up for 2-39 months after surgery,with a median time of 16 months.The postoperative median survival time was 13 months,and 9 patients had tumor recurrence or metastasis during the follow-up.Conclusion The 3D reconstruction virtual surgery planning can accurately complete the preoperative evaluation,meanwhile,it can also provide important reference for the surgical therapy of hilar cholangiocarcinoma.
ABSTRACT
Objective To determine the impact and the risk factors of different methods of preoperative biliary drainage (PBD) for malignant obstruction jaundice (MOJ) on overall survival (OS).Methods Databases including the PubMed,Medline,Web of Knowledge,and other databases were searched up to 30th April,2018 for clinical studies which compared the OS rates between percutaneous transhepatic biliary drainage (PTBD) and endoscopic biliary drainage (EBD) for MOJ.Hazard ratio (HR) and Odds Ratio (OR) with 95％ confidence interval (CI) were performed using the Review Manager 5.3 software to synthesize the results.Results Nine studies were enrolled in this meta-analysis,which included 818 patients in the PTBD group and 1253 patients in the EBD group.EBD was shown to be superior to PTBD in OS (HR=0.63,95％ CI:0.51～0.77,P＜0.05).Risk factors analysis showed that patients in the EBD group had a higher rate of early tumor stage (P＜0.05) and a lower rate of lymphatic metastasis (P＜0.05).When compared with the PTBD group,the EBD group had a lower rate of intraoperative bleeding (P＜0.05),and a higher rate of adjuvant therapy (P＜0.05).Conclusion In PBD for patients with resectable MOJ,there was insufficient evidence to support EBD to be superior to PTBD in OS.
ABSTRACT
Objective To analyze the diagnostic value of serum procalcitonin ( PCT ) for early postoperative bacterial infection after pediatric living donor liver transplantation.Methods A retrospective study was conducted in pediatric patients after living donor liver transplantation recipients admitted to department of critical care medicine of Beijing Friendship Hospital affiliated to Capital Medical University during June 2013 to October 2015.According to the clinical data , all pediatric patients were divided into infection group(n=60) and non-infection group (n=100).Primary disease, PCT post operation day 1 to day 5 for non-infection group and day 1 to day 9 for infection group , temperature , white blood cell , cold ischemia time, warm ischemia time, operation time, volume of blood loss during operation were recorded.All parameters above were compared between groups.Receiver operating characteristic ( ROC) curve was plotted, and the diagnostic value of PCT was evaluated.Results PCT of both groups were elevated after liver transplantation , there was a markedly resolution in non-infection group within 48 to 72 hours.PCT of pediatric patients with bacterial infection was significantly higher than that of non-infected patients , and the difference was of greatly significant (4.62 ±1.39) ng/ml vs (0.85 ±0.19) ng/ml,t=26.56,P=0.00.ROC curve showed that the peak level of PCT might be valuable in the diagnosis of bacterial infection ( AUC=0.985).There was no significant difference of cold ischemia time [(109.92 ±19.22) min vs (108.04 ± 13.20) min, t=1.05, P=0.29], warm ischemia time[(1.49 ±0.17) min vs (1.52 ±0.12) min, t=1.08, P=0.28], operation time[(8.01 ±0.77)vs (8.00 ±1.05) h, t=0.06, P=0.94], WBC[(8.95 ±1.69) ×109/L vs (8.98 ±2.00) ×109/L,t=-0.08, P=0.93]and body temperature[(37.5 ±0.7) vs (37.5 ±0.8) ℃,t=-0.05, P=0.96] on the first day after surgery between infection and non-infection groups.Amount of bleeding in infection group was higher than that of non infection group [ ( 650.87 ± 90.36) ml vs (240.29 ±67.67) ml, t=32.33, P=0.00], there was longer length of ICU stay in the infection group[(11.01 ±1.81)d vs (6.03 ±1.65)d, t=17.78, P=0.00].Conclusion Peak PCT level was a valuable indicator for early postoperative bacterial infection after pediatric living donor liver transplantation.