The best parameters in retinal nerve fiber layer thickness measured by optical coherence tomography in the early stage of glaucoma / 第三军医大学学报

Objective To determine the characteristics and the best parameters of the retinal nerve fiber layer (RNFL) thickness measured by optical coherence tomography (OCT) in the early diagnosis of primary open angle glaucoma (POAG). Methods The thickness of RNFL was quantified by OCT (Carl Zeiss Meditec, Stratus OCT3000 version 3, RNFL3.4) in 71 POAGs(119 eyes) and 76 controls(148 eyes). Results The 9 OCT parameters with the largest AUCs and asymptatic significance less than 0.05 in ROC curve in the early diagnosis of primary open angle glaucoma (POAG) were thickness of 6, 7, 11 o’clock position, inferior, superior, average thickness, inferior maximum, superior average, inferior average. Conclusion In early stage of glaucoma, the thickness of RNFL measured by OCT provides a new index for the diagnosis of glaucoma. The best parameters seem to 6, 7, 11 o’clock position, inferior, superior, average thickess, inferior maximum, superior average, inferior average.

A study of case-mix method by clinical pathway / 医学研究生学报

Objective: To investigate the case-mix method by clinical pathway. Methods: K-MEANS cluster analysis was applied to case-mix classification and artificial neural network was used for case-mix prediction. Results: Five hundred and twenty three inpatient records constructed a case-mix classification scheme of 4 groups.Statistical significant difference of costs existed in 4 groups.The training error of artificial neural network was low(0.0 029) and the predicting result was accurate(98.91%). Conclusion: Case-mix result was more reasonable using records under clinical pathway.The existing models of case-mix depend on dividing individual variables, but artificial neural network does not.

Application of Neural Network Method on the Quantitive Structure-Activity Relationship of Quinolones Compounds / 中国药房

OBJECTIVE:To study the quantitative structure-activity relationship of quinolones compounds by neural network(NN)method.METHODS:A3-layered BP neural network was constructed with the Matlab software package,the collected data were calculated,result of which was compared with that of the linear regression.RESULTS:Sum of square of errors for the neural network method was0.3042,which was less than that of linear regression;the predicted correlation co-efficient was0.86.CONCLUSION:The neural network method has achieved more precise fitting results than the linear re-gression in the study of the quantitative structure-activity relationship of quinolones compounds.