Comparing Three Data Mining Methods to Predict Kidney Transplant Survival

Abstract

Introduction:One of the most important complications of post-transplant is rejection. Analyzing survival is one of the areas of medical prognosis and data mining, as an effective approach, has the capacity of analyzing and estimating outcomes in advance through discovering appropriate models among data. The present study aims at comparing the effectiveness of C5.0 algorithms, neural network and C&RTree to predict kidney transplant survival before transplant.Method:To detect factors effective in predicting transplant survival, information needs analysis was performed via a researcher-made questionnaire. A checklist was prepared and data of 513 kidney disease patient files were extracted from Sina Urology Research Center. Following CRISP methodology for data mining, IBM SPSS Modeler 14.2, C5.0, C&RTree algorithms and neural network were used.Results:Body Mass Index (BMI), cause of renal dysfunction and duration of dialysis were evaluated in all three models as the most effective factors in transplant survival. C5.0 algorithm with the highest validity (96.77%) was the first in estimating kidney transplant survival in patients followed by C&RTree (83.7%) and neural network (79.5%) models.Conclusion:Among the three models, C5.0 algorithm was the top model with high validity that confirms its strength in predicting survival. The most effective kidney transplant survival factors were detected in this study; therefore, duration of transplant survival (year) can be determined considering the regulations set for a new sample with specific characteristics.

Promises of Big Data and Artificial Intelligence in Nephrology and Transplantation

“…In order to reinforce the usage and subsequent transformation of AI as well as data-based CDSSs in nephrology, AI, as well as big data, offers the chance to actually source knowledge from expert knowledge and big data and subsequently transform it into some form of intelligent system, which can be applied in risk classification, disease diagnosis, drug discovery, and prognostic evaluation, among some other things. AI might be useful in establishing the type of kidney disease and subsequently help in solving problems related to survival analysis of the patients who have gone through kidney transplants [106][107][108][109][110][111][112][113][114]. Renal biopsy images may be a good data base for application of machine learning algorithms.…”

Application of Artificial Intelligence Techniques to Predict Survival in Kidney Transplantation: A Review

“…To do so, the two most highly ranked factors in each of the papers analysed have been selected (two have been selected because Shahmoradi et al, 2016 only has two). Five out of the eight papers mention the factors that influence survival [7,31,32,[35][36][37]. Other factors influencing survival that was mentioned in the papers review are: hypertension, smoking, a history of viral hepatitis B and C, cerebral and peripheral vascular disease, recipient ethnicity category or recipient HCV status, among others.…”

“…A more detailed explanation of the decision trees can be found in Appendix A. According to the review carried out, in the survival analysis, there are several types of decision trees used like CART (Classification And Regression Trees) [40][41][42][43][44], C5.0 [36], J48 [34] and the survival decision tree model. CART [45] is a method that works with all types of variables, with no need to make continuous features discrete, so that the classification is used when the variable of interest is categorical and regression is used in the case of a continuous variable.…”

“…The authors in [36] evaluated the C5.0, multilayer perceptron of artificial neural network, and CART to predict kidney transplant survival before the transplant. The authors do not mention any specific treatment for censored data.…”

Developing an Intelligent System for Prediction of Optimal Dose of Warfarin in Iranian Adult Patients with Artificial Heart Valve

“…Many clinical and genetic features affect the dosage of patients [5,9,11] therefore, machine learning techniques can be used for the diagnosis or prediction of warfarin's doses to help expert humans to make the best decisions [12,13]. Artificial Neural Networks (ANNs) are one of the most important ML algorithms [14] which are apt for such applications as: first, they can incorporate large number of inputs and large number of outputs; second, they are dynamic and can be upgraded with new data sets periodically; third, they can explore the nonlinear relationship; forth, they have high generalizability [15,16] and hence, can be used to predict warfarin dosage requirements.…”