Forecasting The Risk Of Chronic Kidney Disease Using Artificial Intelligence Approaches

In this Paper With the aid of AI techniques, this study aims to predict the early detection of chronic kidney disease, also known as chronic renal disease, in diabetic patients. It then suggests a decision tree to reach specific conclusions with desired accuracy by evaluating its performance in relation to its specification and sensitivity. Methods: The behaviour of learning algorithms based on a set of data mining indicators affects the models that are produced proportionately. Predicting the future is no longer a difficult task thanks to the promises of predictive analytics in big data and the use of machine learning algorithms, especially for the health sector, which has undergone significant evolution as a result of the development of new computer technologies that gave rise to numerous fields of study research. Many initiatives are made to deal with the explosion of medical data on the one hand, and to learn meaningful information from it, forecast diseases, and anticipate treatments on the other. To extract meaningful information and aid in decision-making, researchers used all the technological advancements, including big data analytics, predictive analytics, machine learning, and learning algorithms.

Application of routine laboratory data in combination with machine learning in the differential diagnosis of lung tuberculosis / 中华检验医学杂志

Objective:To investigate the application value of establishing the differential diagnosis model of pulmonary tuberculosis using routine laboratory data.Methods:The retrospective study was conducted. The routine laboratory data of newly diagnosed patients with pulmonary tuberculosis and other pulmonary diseases in Beijng Jishuitan Hospital and Beijing Hepingli Hospital from May 2015 to November 2021were collected. According to the random numbers showed in the computer, all the 11516 patients were divided into training dataset and test dataset with a ratio of 9∶1. Four machine learning algorithms, Support Vector Machine, Random Forest, K-Nearest Neighbor and Logistic Regression, were used to build models and select features. The diagnostic accuracy of each model was verified by using the 10-fold cross-validation method and the performance of each model was evaluated by using the receptor operator of characteristic (ROC) curve.Results:Random Forest was selected as the optimal machine learning algorithm to build the best feature model in the study. According to importance scale of factors, the differential diagnosis model of pulmonary tuberculosis consisting of 37 non-specific test indexes. In the validation set and test set the accuracy and area under curve (AUC) of the models were 0.747 and 0.736, the sensitivity, specificity and accuracy were 68.03% and 68.75%, 70.91% and 67.90%, 70.30% and 68.12%, respectively.Conclusion:A key tool in the differential diagnosis model of pulmonary tuberculosis was established by routine laboratory data in combination with machine learning. The results of this study need to be further verified by more data from medical institutions.

Effect of intelligent early warning scoring system in condition monitoring of hospitalized non-critically ill patients / 中华急诊医学杂志

Objective:To establish and apply the electronic further modified early warning score system (e-fMEWS), and explore its role in the condition evaluation and early warning of inpatients in non-critical units, so as to provide clinical nurses with an early and dynamic method to identify the potential deterioration risk of patients' condition.Methods:A retrospective analysis of 262 805 inpatients in multiple non-critical units of the Second Affiliated Hospital of Zhejiang University School of Medicine from January to December 2018 and January to December 2020 was performed. The patients who were hospitalized from January to December 2018 were used as the control group, and the responsible nurse used the traditional single evaluation index to start the emergency response system; the patients from January to December 2020 were used as the research group, and the emergency response system was started using e-fMEWS. The inclusion criteria were as follows: (1) hospitalization time ≥24 h; (2) patient ≥14 years old. Exclusion criteria were as follows: (1) patients had cardiopulmonary resuscitation before admission; (2) patients discontinued treatment or were transferred to another hospital during treatment; (3) patients received palliative care; (4) patients were admitted to non-critical wards in grade I of emergency pre-examination and triage. The activation of the rapid response team (RRT), the activation of the cardiorespiratory arrest team, the incidence of cardiac and respiratory arrest, the number of cases of invasive mechanical ventilation, the number of cases admitted to the intensive care unit, the length of hospital stay and the prognosis were compared. Statistical software SPSS 22.0 was used for data analysis.Results:Under the e-fMEWS assessment, compared with the control group, the rate of initiation of the research group decreased by 0.03%. For patients who initiated RRT, the average length of hospital stay was shortened, and the number of in-hospital respiratory cardiac arrest decreased (12.2% vs. 13.2%) and the number of cases transferred to the intensive care unit was less (42.8% vs. 50.6%), the rate of improvement and recovary increased (58.4% vs. 56.1%).Conclusions:The application of e-fMEWS can help clinical nurses to quickly and accurately identify the potential risk of deterioration of the patient's condition. Through early identification of potentially critically ill patients in non-critical units, early intervention and timely treatment can avoid adverse events and improve the patient prognosis.

The development of a decision support system for diagnosing nasal allergy

This paper deals with the problem of improving the capability of the medical decision support system (MDSS) for diagnosing nasal allergy by integrating the previously developed expert system with the neural network approach. Three knowledge acquisition methods were used to develop the expert system: statistical, rule-based, and the combined approach. Among the three, a combined approach showed the best prediction rate based on discriminant analysis. Using the results of a combined approach as input values, the neural network was developed using back-propagation method. Unlike the expert system, the neural network system provides the resulting allergy status in probabilistic terms. Managerial as well as legal issues were also discussed in this paper.