ABSTRACT
The application of artificial neural network algorithm in pathological diagnosis of gastrointestinal malignant tumors has become a research hotspot. In the previous studies, the algorithm research mainly focused on the model development based on convolutional neural networks, while only a few studies used the combination of convolutional neural networks and recurrent neural networks. The research contents included classical histopathological diagnosis and molecular typing of malignant tumors, and the prediction of patient prognosis by utilizing artificial neural networks. This article reviews the research progress on artificial neural network algorithm in the pathological diagnosis and prognosis prediction of digestive tract malignant tumors.
Subject(s)
Humans , Neural Networks, Computer , Algorithms , Prognosis , Gastrointestinal Neoplasms/diagnosisABSTRACT
Objective:To analyze the etiology of biliary fungal infection and risk factors of case fatality.Methods:Clinical and laboratory data of 91 biliary fungal infection patients admitted in Li Huili Hospital of Ningbo Medical Center from January 2013 to June 2019 were retrospectively reviewed, including 14 patients (16.4%) with fungal infection and 77 patients (84.6%) with fungal and bacterial mixed infection. There were 79 survivors and 12 deaths, the risk factors of fatality were analyzed by binary Logistic regression analysis.Results:The fungal strain Candida albicans was detected in 61 cases, Candida glabrata in 19 cases and Candida tropicalis in 6 cases. Drug sensitivity test showed that the fungal strains were highly sensitivity to amphotericin B and 5-fluorouracil [100.0%(91/91),97.8%(89/91)]. In 77 mixed infection cases Gram-negative bacteria was the more common (34 cases,44.2%). The average age of patients was 70.7 years old. Benign diseases were found in 66 cases (72.5%) and 61(67.0%)of them were cholelithiasis. Patients with a history of repeated biliary operation were more likely to have mixed infection of biliary fungi and bacteria (χ 2= 4.56, P=0.03). The mean albumin level in the fatal group was significantly lower than that in the survival group [(28.1±5.2)g/L vs. (33.3±5.3)g/L; t=2.77, P=0.01]. The median length of hospital stay in the survival group was significantly shorter than that in the fatal group [12.0(9.0, 18.0)d vs. 29.5 (13.0, 42.7)d; Z=-2.37, P=0.02]. Multiple logistic regression analysis showed that the history of repeated biliary operation ( OR=4.46, 95% CI: 1.06—4.97) and mixed infection of fungi with bacteria ( OR=10.20, 95% CI: 1.48—70.27) were the risk factors of case fatality. Conclusion:Candida albicans is the main fungus in biliary fugal infection which is often complicated with bacterial infection. Repeated biliary operations and mixed infection of fungi with bacteria are the risk factors of death in patients with biliary infection.
ABSTRACT
Objective:To explore the predictive value of 18F-fluorodeoxyglucose (FDG) PET/CT radiomics for the programmed death ligand-1 (PD-L1) expression level in lung adenocarcinoma patients. Methods:A total of 101 patients (43 males, 58 females; median age 60 years) with histologically confirmed lung adenocarcinoma who received pre-treatment 18F-FDG PET/CT from January 2017 to January 2019 in Peking University Cancer Hospital were included retrospectively. There were 44 patients with positive PD-L1 by immunohistochemical assays, and 57 with PD-L1 negative. Patients were assigned to a training set ( n=71) and a validation set ( n=30). Clinical data, PET/CT radiomics parameters, conventional metabolic parameters, and observed CT characteristics of these patients were included in the models. The filter method and embedded method were used in feature selection. Models based on logistic regression, random forest, XGBoost and Light Gradient Boosting Machine (LightGBM) were trained and evaluated, and the optimal parameters to predict the PD-L1 expression as well as the area under curve (AUC) were attained. Results:All models had predictive ability in the prediction of PD-L1 expression, while LightGBM was more powerful than the others, with the precision for positive and negative predictions of 0.85 and 0.76, respectively. Incorporating clinical data and data derived from thin-section CT images (clinical data+ CT) into the LightGBM, the precision, recall and F1-score for positive and negative patients were 0.71, 0.67, 0.69 and 0.69, 0.73, 0.72, respectively, with the accuracy of 0.70 and the AUC of 0.79. As for clinical data+ PET, the precision, recall and F1-score for positive and negative patients were 0.79, 0.73, 0.76 and 0.75, 0.80, 0.77, respectively, with the accuracy of 0.77 and the AUC of 0.80. As for clinical data+ CT+ PET, the precision, recall and F1-score for positive and negative patients were 0.85, 0.73, 0.79 and 0.76, 0.87, 0.81, respectively, with the accuracy of 0.80 and the AUC of 0.83. Features with significant importance in the model (clinical data+ CT+ PET) were as follows: maximum standardized uptake value (SUV max), peak of standardized uptake value (SUV peak), CT_shape_Maximum2DDiameterSlice, PET_shape_Elongation, PET_gray level co-occurrence matrix (GLCM)_Correlation, etc. Conclusions:Incorporating clinical data, PET/CT radiomics features and conventional metabolic parameters, the PD-L1 expression can be effectively predicted, which help to assist the selection of patients who may benefit from the immunotherapy.
ABSTRACT
Background:Gastric cancer is the second most frequently diagnosed cancer and the third cause of cancer death in China. Selection of candidates in high risk of gastric cancer by a simple and non-invasive marker with high sensitivity and then undergoing endoscopy is an optimal approach for large scale gastric cancer screening. Aims:To evaluate the clinical value of using trefoil factor 3(TFF3)as a serum biomarker for gastric cancer screening. Methods:Serum samples of 49 gastric cancer patients and 29 healthy subjects were collected for measurements of serum TFFs by ELISA from Jul. 2013 to Jan. 2014 at Ningbo Medical Treatment Center Lihuili Hospital. ROC curve and the area under curve(AUC)were used to verify the diagnostic performance of serum TFF1,TFF2 and TFF3 for gastric cancer. Correlations between TFFs and clinicopathological characteristics of gastric cancer were further analyzed. Results:Serum concentration of TFF3 in gastric cancer group was significantly higher than that in healthy controls[(43. 57 ± 19. 49)ng/ mL vs.(29. 97 ± 14. 20)ng/ mL, P 0. 05). AUC of serum TFF1,TFF2 and TFF3 for diagnosis of gastric cancer were 0. 56,0. 56 and 0. 83, respectively,which indicated that the performance of TFF3 was the best. Taken 33. 0 ng/ mL as the cut off value of TFF3, the sensitivity and specificity were 63. 3% and 82. 8% ,respectively,and the odds ratio for predicting gastric cancer was 8. 27. Significant correlations were existed between serum concentration of TFF3 and TNM stage,differentiation and lymph node metastasis of gastric cancer(P < 0. 05). Conclusions:Serum TFF3 is a promising non-invasive biomarker for gastric cancer screening.
ABSTRACT
Establishment of quality management system (QMS) plays a critical role in the clinical data management (CDM). The objectives of CDM are to ensure the quality and integrity of the trial data. Thus, every stage or element that may impact the quality outcomes of clinical studies should be in the controlled manner, which is referred to the full life cycle of CDM associated with the data collection, handling and statistical analysis of trial data. Based on the QMS, this paper provides consensus on how to develop a compliant clinical data management plan (CDMP). According to the essential requirements of the CDM, the CDMP should encompass each process of data collection, data capture and cleaning, medical coding, data verification and reconciliation, database monitoring and management, external data transmission and integration, data documentation and data quality assurance and so on. Creating and following up data management plan in each designed data management steps, dynamically record systems used, actions taken, parties involved will build and confirm regulated data management processes, standard operational procedures and effective quality metrics in all data management activities. CDMP is one of most important data management documents that is the solid foundation for clinical data quality.
ABSTRACT
Objective:To investigate the diagnostic value of the promoter methylation of plasma RNF180 gene and its protein ex-pression for the detection of gastric cancer. Methods:Methylation-specific polymerase-chain reaction (MSP) and enzyme-linked immu-no-sorbent assay (ELISA) were performed to detect DNA methylation and protein expression of the RNF180 gene, respectively. The correlations of DNA methylation and protein expression of the RNF180 gene with the clinico-pathological parameters of gastric carcino-ma were then separately analyzed. Results:MSP showed that the methylation rates of the RNF180 gene were 62.75%and 21.88%in the plasma of patients with gastric carcinoma and healthy volunteers, respectively;this result indicated that the two groups significantly differed (P0.05). Conclusion:The RNF180 gene is expressed at a hypermethylation rate, and the corresponding protein expression level is de-creased in the plasma of individuals with gastric carcinoma. Therefore, RNF180 gene methylation in plasma could be applied to detect microinvasion for the clinical diagnosis of gastric cancer.