Objective To test the hypothesis that the macular pigment may be a marker of foveal cone function and consequently the structural integrity of foveal cones.Methods Sixteen patients (32 eyes) diagnosed to have Stargardt dystrophy and three patients with full thickness macular holes by clinical criteria were studied with a scanning laser ophthalmoscopy (SLO) comparing argon laser blue and infrared images for the presence or absence of macular pigment (MP) in the fovea. An C++ computer based program was used to evaluate the density of MP. Eyes were graded into three categories: those without foveal macular pigment, those with partial pigment and those with normal amounts of macular pigment. These categories were compared with visual acuity determined by the Snellen chart. Results Thirteen eyes with a visual acuity of 20/200 or worse had no macular pigment in the fovea. Eleven eyes with visual acuity of 20/40 or better had a normal amount of macular pigment in the fovea and 1 eye had partial macular pigment. Eleven eyes with partial macular pigment had intermediary acuity value.Conclusions Foveal macular pigment is closely related to foveal cone acuity and therefore may be a marker for the presence of foveal cones. Infrared light is a sensitive indicator of early macular diseases.(Chin J Ocul Fundus Dis，2003，19：201-268)
Evidence is the core of Evidence-Based Medcine; the Grades of Recommendations Assessment, Development and Evaluation (GRADE System) is a milestone in the history of evidence development. This paper outlines the GRADE System and GRADEpro 3.2 software, and briefly explores the right and wrong application which was published in the Chinese Journal of Evidence-Based Medicine. The GRADEpro 3.2 software is easy to operate, but for evaluating the reasons of upgrade and downgrade, and the importance of the parameters of outcomes, it needs to comprehensively and systematically understand the knowledge of relevant background, and to construct a solid foundation in clinical epidemiology and systematic review. In view of this paper based on the current GRADE System, there may be some discrepancy to the later content with the GRADE System constant improvement. Therefore, it is bly recommended that readers should keep constant learning and improving.
There have been problems in the existing multiple physiological parameter real-time monitoring system, such as insufficient server capacity for physiological data storage and analysis so that data consistency can not be guaranteed, poor performance in real-time, and other issues caused by the growing scale of data. We therefore proposed a new solution which was with multiple physiological parameters and could calculate clustered background data storage and processing based on cloud computing. Through our studies, a batch processing for longitudinal analysis of patients' historical data was introduced. The process included the resource virtualization of IaaS layer for cloud platform, the construction of real-time computing platform of PaaS layer, the reception and analysis of data stream of SaaS layer, and the bottleneck problem of multi-parameter data transmission, etc. The results were to achieve in real-time physiological information transmission, storage and analysis of a large amount of data. The simulation test results showed that the remote multiple physiological parameter monitoring system based on cloud platform had obvious advantages in processing time and load balancing over the traditional server model. This architecture solved the problems including long turnaround time, poor performance of real-time analysis, lack of extensibility and other issues, which exist in the traditional remote medical services. Technical support was provided in order to facilitate a "wearable wireless sensor plus mobile wireless transmission plus cloud computing service" mode moving towards home health monitoring for multiple physiological parameter wireless monitoring.
The accuracy of the clinical prediction model determines its extrapolation and application value. When the prediction model is applied to a new setting, the differences between the new population and the initially modeled population in terms of study time, population characteristics, region, and other factors could lead to a reduction in its predictive performance. Calibrating or updating the prediction model with appropriate statistical methods is important to improve the accuracy of the prediction model in new populations. The model updating methods mainly include regression coefficients updating, meta-model updating and dynamic model updating. However, due to the limitations of meta-model updating and dynamic model updating in practical applications, the regression coefficient updating method is still the most common method in model updating. This paper introducd several types of model updating methods, the regression coefficients updating methods for two common clinical prediction models based on Logistic regression and Cox regression, and provide corresponding R codes for reference of researchers.