A proof of concept reinforcement learning based tool for non parametric population pharmacokinetics workflow optimization.

The building of population pharmacokinetic models can be described as an iterative process in which given a model and a dataset, the pharmacometrician introduces some changes to the model specification, then perform an evaluation and based on the predictions obtained performs further optimization. This process (perform an action, witness a result, optimize your knowledge) is a perfect scenario for the implementation of Reinforcement Learning algorithms. In this paper we present the conceptual background and a implementation of one of those algorithms aiming to show pharmacometricians how to automate (to a certain point) the iterative model building process.We present the selected discretization for the action and the state space. SARSA (State-Action-Reward-State-Action) was selected as the RL algorithm to use, configured with a window of 1000 episodes with and a limit of 30 actions per episode. SARSA was configured to control an interface to the Non-Parametric Optimal Design algorithm, that was actually performing the parameter optimization.The Reinforcement Learning (RL) based agent managed to obtain the same likelihood and number of support points, with a distribution similar to the reported in the original paper. The total amount of time used by the train the agent was 5.5 h although we think this time can be further improved. It is possible to automatically find the structural model that maximizes the final likelihood for an specific pharmacokinetic dataset by using RL algorithm. The framework provided could allow the integration of even more actions i.e: add/remove covariates, non-linear compartments or the execution of secondary analysis. Many limitations were found while performing this study but we hope to address them all in future studies.

Misunderstandings and general strategies of statistical modeling in medical research papers / 中华医学科研管理杂志

Objective To summarize and analyze the Logistic and linear regression modeling of medical research papers in 2018,to propose a general modeling strategy.Methods Search 2018 China Knowledge Network medical research related papers,extract some papers for evaluation,identify and analyze possible modeling defects in the paper writing process,provide the general method of modeling.Results In the China Knowledge Network database,1 319 medical research papers were detected in 2018,and 125 papers were randomly selected for evaluation.Identified issues include no data cleaning before modeling,insufficient attention to modeling,and model evaluation after modeling.Conclusions There are defects in the modeling process of medical research papers,and further attention and enhancement are needed in the writing process.

Integración de técnicas de Minería de procesos para la detección de la variabilidad en procesos hospitalarios desde sistemas automatizados / Integrating mining processes techniques in hospital procedures for the detection of variability from their information systems

Los sistemas de información hospitalaria cuentan con un volumen importante de datos, sin embargo, carecen de mecanismos que permitan analizar la ejecución de los procesos e identificar variabilidad. La variabilidad puede observarse en prácticamente cada paso del proceso asistencial y a varios niveles de agrupación: poblacional e individual. Desde el punto de vista poblacional se comparan tasas de realización de un procedimiento clínico, como pueden ser intervenciones quirúrgicas o ingresos hospitalarios en un período de tiempo. Las técnicas de minería de procesos analizan los datos reales de sistemas informáticos y son útiles para la detección de variabilidad en la ejecución de los procesos de negocio. La presente investigación propone la aplicación de técnicas de minería de procesos, seleccionadas a partir de un riguroso estudio del estado del arte, para el análisis de los procesos hospitalarios desde sus sistemas de información y materializadas en un modelo computacional. El Modelo para la Detección de Variabilidad (MDV) se instrumentó exitosamente en el sistema XAVIA HIS desarrollado por la Universidad de las Ciencias Informáticas UCI, donde fueron adaptadas e integradas las técnicas de minería de procesos. El modelo MDV contribuye al proceso de informatización de la salud en Cuba. La solución propicia la utilización de una tecnología emergente en áreas como la industrial y empresarial en el entorno sanitario. Esta beneficia importantes funciones gerenciales como la gestión, control y planificación de recursos y servicios sanitarios(AU)

The hospital information systems collect an important volume of data, however, they lack mechanisms to analyze the execution of the processes and identify variability. In practically every step of the care process and at various levels of grouping: population and individual the variability is present. From a population point of view, performance rates of a clinical procedure such as surgical interventions or hospital admissions, are compared over time. Process mining techniques analyze the real data of computer systems and are useful for the detection of variability in the execution of business processes. Based on a rigorous study of the state of the art, this research proposes the application of process mining techniques for the analysis of hospital processes from their information systems, providing a computational model. Model for Variability Detection (MDV) implemented successfully in the XAVIA HIS system developed by the UCI University of Informatics Sciences, where techniques of process mining were adapted and integrated. The MDV model contributes to the process of computerization of health in Cuba. The solution encourages the use of an emerging technology in areas such as industrial and business in the healthcare environment. This benefits important management functions such as control and planning of resources and health services(AU)

Integración de técnicas de Minería de procesos para la detección de la variabilidad en procesos hospitalarios desde sistemas automatizados / Integrating mining processes techniques in hospital procedures for the detection of variability from their information systems

Los sistemas de información hospitalaria cuentan con un volumen importante de datos, sin embargo, carecen de mecanismos que permitan analizar la ejecución de los procesos e identificar variabilidad. La variabilidad puede observarse en prácticamente cada paso del proceso asistencial y a varios niveles de agrupación: poblacional e individual. Desde el punto de vista poblacional se comparan tasas de realización de un procedimiento clínico, como pueden ser intervenciones quirúrgicas o ingresos hospitalarios en un período de tiempo. Las técnicas de minería de procesos analizan los datos reales de sistemas informáticos y son útiles para la detección de variabilidad en la ejecución de los procesos de negocio. La presente investigación propone la aplicación de técnicas de minería de procesos, seleccionadas a partir de un riguroso estudio del estado del arte, para el análisis de los procesos hospitalarios desde sus sistemas de información y materializadas en un modelo computacional. El Modelo para la Detección de Variabilidad (MDV) se instrumentó exitosamente en el sistema XAVIA HIS desarrollado por la Universidad de las Ciencias Informáticas UCI, donde fueron adaptadas e integradas las técnicas de minería de procesos. El modelo MDV contribuye al proceso de informatización de la salud en Cuba. La solución propicia la utilización de una tecnología emergente en áreas como la industrial y empresarial en el entorno sanitario. Esta beneficia importantes funciones gerenciales como la gestión, control y planificación de recursos y servicios sanitarios(AU)

The hospital information systems collect an important volume of data, however, they lack mechanisms to analyze the execution of the processes and identify variability. In practically every step of the care process and at various levels of grouping: population and individual the variability is present. From a population point of view, performance rates of a clinical procedure such as surgical interventions or hospital admissions, are compared over time. Process mining techniques analyze the real data of computer systems and are useful for the detection of variability in the execution of business processes. Based on a rigorous study of the state of the art, this research proposes the application of process mining techniques for the analysis of hospital processes from their information systems, providing a computational model. Model for Variability Detection (MDV) implemented successfully in the XAVIA HIS system developed by the UCI University of Informatics Sciences, where techniques of process mining were adapted and integrated. The MDV model contributes to the process of computerization of health in Cuba. The solution encourages the use of an emerging technology in areas such as industrial and business in the healthcare environment. This benefits important management functions such as control and planning of resources and health services(AU)

Modeling number of bacteria per food unit in comparison to bacterial concentration in quantitative risk assessment: impact on risk estimates.

When developing quantitative risk assessment models, a fundamental consideration for risk assessors is to decide whether to evaluate changes in bacterial levels in terms of concentrations or in terms of bacterial numbers. Although modeling bacteria in terms of integer numbers may be regarded as a more intuitive and rigorous choice, modeling bacterial concentrations is more popular as it is generally less mathematically complex. We tested three different modeling approaches in a simulation study. The first approach considered bacterial concentrations; the second considered the number of bacteria in contaminated units, and the third considered the expected number of bacteria in contaminated units. Simulation results indicate that modeling concentrations tends to overestimate risk compared to modeling the number of bacteria. A sensitivity analysis using a regression tree suggests that processes which include drastic scenarios consisting of combinations of large bacterial inactivation followed by large bacterial growth frequently lead to a >10-fold overestimation of the average risk when modeling concentrations as opposed to bacterial numbers. Alternatively, the approach of modeling the expected number of bacteria in positive units generates results similar to the second method and is easier to use, thus potentially representing a promising compromise.

Study on the Preparing and Quality Control for Golden Gel / 世界科学技术-中医药现代化

This article is aimed to study the optimal process of preparation of Golden water-soluble gel, and to establish its quality control method. Taking the gel forming properties, stability, water loss rate, pH as investigation index, the best extraction process was screened by single factor and orthogonal test design. TLC method was used to identify the product of trichosanthin, rhubarb, and the content of paeoniflorin was determined by HPLC method. The optimal condition of preparation process is to use 1% carbomer as gel matrix, 10% glycerol as humectants, and 0.5%triethanolamine as pH modulators; trichosanthes and rhubarb could be detected by TLC method, with pH range 5.00 - 5.15. The linear range of paeoniflorin was 29.04-945.20μg·mL-1, the average recovery was 98.80%, and RSD was 2.91% (n=9). Golden gel forming process is simple and quality controllable with reliability and good stability.

Estimation of Jaw and MLC Transmission Factor Obtained by the Auto-modeling Process in the Pinnacle3 Treatment Planning System / 의학물리

Radiation treatment techniques using photon beam such as three-dimensional conformal radiation therapy (3D-CRT) as well as intensity modulated radiotherapy treatment (IMRT) demand accurate dose calculation in order to increase target coverage and spare healthy tissue. Both jaw collimator and multi-leaf collimators (MLCs) for photon beams have been used to achieve such goals. In the Pinnacle3 treatment planning system (TPS), which we are using in our clinics, a set of model parameters like jaw collimator transmission factor (JTF) and MLC transmission factor (MLCTF) are determined from the measured data because it is using a model-based photon dose algorithm. However, model parameters obtained by this auto-modeling process can be different from those by direct measurement, which can have a dosimetric effect on the dose distribution. In this paper we estimated JTF and MLCTF obtained by the auto-modeling process in the Pinnacle3 TPS. At first, we obtained JTF and MLCTF by direct measurement, which were the ratio of the output at the reference depth under the closed jaw collimator (MLCs for MLCTF) to that at the same depth with the field size 10x10 cm2 in the water phantom. And then JTF and MLCTF were also obtained by auto-modeling process. And we evaluated the dose difference through phantom and patient study in the 3D-CRT plan. For direct measurement, JTF was 0.001966 for 6 MV and 0.002971 for 10 MV, and MLCTF was 0.01657 for 6 MV and 0.01925 for 10 MV. On the other hand, for auto-modeling process, JTF was 0.001983 for 6 MV and 0.010431 for 10 MV, and MLCTF was 0.00188 for 6 MV and 0.00453 for 10 MV. JTF and MLCTF by direct measurement were very different from those by auto-modeling process and even more reasonable considering each beam quality of 6 MV and 10 MV. These different parameters affect the dose in the low-dose region. Since the wrong estimation of JTF and MLCTF can lead some dosimetric error, comparison of direct measurement and auto-modeling of JTF and MLCTF would be helpful during the beam commissioning.

Adecuación de la definición de caso sospechoso de dengue basado en la asociación de síntomas y signos según los registros médicos, Costa Rica, 1998 / Adjustment of definition of suspicious case of dengue based on the association of symptoms and signs according to the medical registries, Costa Rica, 1998

A partir de octubre de 1993, en Costa Rica se estableció una definición de caso sospechoso de dengue, basada en la presencia de fiebre y dos o más de los siguientes signos y síntomas: cefalea, mialgias, artralgias, dolor retroocular y exantema, la cual, aplicada al sistema de vigilancia, le confiere una sensibilidad general de 36%. Con el propósito de sugerir una definición de caso sospechoso de dengue que pueda mejorar la sensibilidad del sistema, se estudió la probabilidad de que el personal médico diagnostique a un paciente con dengue está o no relacionada con la presencia de ciertos síntomas y signos.Para ello, se realizó un estudio observacional de tipo analítico para someter a prueba, mediante una regresión logística, los diferentes síntomas y signos que podrían estar asociados a la infección por el virus dengue. Se consideraron los criterios clínicos anotados por el personal médico en el expediente y los resultados del análisis del laboratorio de referencia de dengue. La población en estudio correspondió a la totalidad de los expedientes incluidos en el estudio "Evaluación del Sistema de Vigilancia Epidemiológico del Dengue utilizando como indicador la aplicación de la definición de caso sospechoso, Costa Rica, 1998".El mejor modelo estructurado de la definición de caso sospechoso incluyó fiebre (OR=1,7 IC95% 0,7-3,9), mialgias (OR=1,7 IC95% 1-2,9), artralgias (OR=1,6 IC95% 0,8-2,9), exantema (OR=2,8 IC95% 1,3-5,9), dolor retroocular (OR=2,8 IC95% 1,5-5,3) y manifestaciones de sangrado (OR=6,4 IC95% 1,3-30) por ser síntomas y signos que al ser discriminantes, automáticamente desencadenaron un primer diagnóstico diferencial del dengue. Sin embargo, su aplicación al sistema de vigilancia del dengue, le confiere apenas una sensibilidad del 25%...

In October 1993, a case definition for dengue based on the presence of fever and two or more of the following symptoms: headache, mialgia, arthralgia, retroocular pain and exantema, was established in Costa Rica. When applied in the National Surveillance Program, this definition had showed a general sensitivity of 36%. In the present investigation, the probability for the health personnel to diagnose a case of dengue based on the presence of a combination of symptoms and signs, was evaluated. The study was intended to suggest a dengue case definition to improve the sensitivity of the surveillance system. The study population fully correspond to the clinical files included in a previous report ("Evaluación del Sistema de Vigilancia Epidemiológico del Dengue utilizando como indicador la aplicación de la definición de caso sospechoso, Costa Rica, 1998"). This research corresponds to an analytical observational study. The association of difterent symptoms and signs to the possibiíity of diagnosing a dengue virus infection was evaluated by using a logistic regression analysis. Data evaluated included the clinical criteria described by health personnel in the patient's clinical files and the results of the laboratory tests indicated for the confirmation of dengue infection. The results showed that the best structured model for the suspicious case definition included fever (OR=1,7 IC95% 0,7-3,9), myalgia (OR=1,7 IC95% 1-2,9), arthralgia (OR=1,6 IC95% 0,8-2,9), exanthema (OR=2,8 IC95% 1,3-5,9), retroocular pain (OR=2,8 1C95% 1,5-5,3) and bleeding manifestations (OR=6,4 IC95% 1,3-30). The combination of these symptoms and signs allow the discrimination of dengue cases in the first step of the differential diagnosis. However, the application of this combination gives the system a sensitivity of only 25 %...