Machine Learning Models for Predicting the Type and Outcome of Ureteral Stones Treatments.

Background: Kidney and ureter stones are the third pathologies in urological diseases. Less invasive treatments such as transureteral lithotripsy and extracorporeal shock wave lithotripsy are used to treat ureteral stones. Data mining has provided the possibility of improving decision-making in choosing the optimal treatment. In this paper predictive models for the detection of ureter stone treatment (first model) and its outcome (second model) is developed based on the patient's demographic, clinical, and laboratory factors. Methods and Material: In this cross-sectional study a questionnaire was used to identify the most effective features in the predictive models, and Information on 440 patients was collected. The models were constructed using machine learning techniques (Multilayer perceptron, Classification, and regression tree, k-nearest neighbors, Support vector machine, Naïve Bayes classifier, Random Forest, and AdaBoost) in the Bigpro1 analytical system. Results: Among the Holdout and K-fold cross-validation methods used, the Holdout method showed better performance. From the data-based balancing methods used in the second model, the Synthetic Minority oversampling technique showed better performance. Also, the AdaBoost algorithm had the best performance. In this algorithm, accuracy, sensitivity, specificity, precision, F- measure, and Area under the carve in the first model were 89%, 87%, 91%, 90%, 89%, and 94% respectively, and in the second model were 81%, 81%, 82%, 84%, 82%, and 85% respectively. Conclusions: The results were promising and showed that the data mining techniques could be a powerful assistant for urologists to predict a surgical outcome and also to choose an appropriate surgical treatment for removing ureter stones.

Design and evaluation of an educational mobile program for liver transplant patients.

BACKGROUND: Liver transplantation, the last treatment for advanced liver failure, necessitates patient education due to its wide range of complications and subsequent disabilities. The present study was development-applied research and aimed to design a mobile-based educational program to provide liver transplant patients with critical health information. METHODS: In the first phase of the study, the crucial educational components were collected from the literature and organized in the form of a questionnaire using library studies and available global guidelines. The validity and reliability of this researcher-made questionnaire were confirmed by a panel of experts (n = 15), including gastroenterologists and liver specialists working in the Motahari liver clinic and AbuAli Sina Hospital in Shiraz. The application was designed followed by analyzing the data gathered from the first phase. To evaluate the mobile phone program's usability, to evaluate the application, 30 liver transplant patients were randomly selected. RESULTS: Most educational components covered in the questionnaire were deemed necessary by experts in the first phase. As a result, the educational contents were classified under 10 categories. The application had a good level of usability since the participants' satisfaction score was 8.1 (out of 9 points). CONCLUSIONS: Due to the increase in liver transplantation and the use of mobile phones, applications increase the patient's role in their health, and their awareness. It also leads to a better interaction and follow-up of the patient, the treatment staff of the medical centers.

A Machine Learning Model for Detection of Coronary Artery Disease Using Noninvasive Clinical Parameters.

Background and Objective: Coronary artery disease (CAD) is one of the most prevalent causes of death worldwide. The early diagnosis and timely medical care of cardiovascular patients can greatly prevent death and reduce the cost of treatments associated with CAD. In this study, we attempt to prepare a new model for early CAD diagnosis. The proposed model can diagnose CAD based on clinical data and without the use of an invasive procedure. Methods: In this paper, machine-learning (ML) techniques were used for the early detection of CAD, which were applied to a CAD dataset known as Z-Alizadeh Sani. Since this dataset has 54 features, the Pearson correlation feature selection method was conducted to identify the most effective features. Then, six machine learning techniques including decision tree, deep learning, logistic regression, random forest, support vector machine (SVM), and Xgboost were employed based on a semi-random-partitioning framework. Result: Applying Pearson feature selection to the dataset demonstrated that only eight features were the most effective for CAD diagnosis. The results of running the six machine-learning models on the selected features showed that logistic regression and SVM had the same performance with 95.45% accuracy, 95.91% sensitivity, 91.66% specificity, and a 96.90% F1 score. In addition, the ROC curve indicates a similar result regarding the AUC (0.98). Conclusions: Prediction is an important component of medical decision support systems. The results of the present study showed that feature selection has a high impact on machine-learning performance and, regardless of the evaluation metrics of the machine-learning models, determining the effective features is very important. However, SVM and Logistic Regression were designated as the best models according to our selected features.