Comparison and development of machine learning tools in the prediction of chronic kidney disease progression

Abstract

Background Urinary protein quantification is critical for assessing the severity of chronic kidney disease (CKD). However, the current procedure for determining the severity of CKD is completed through evaluating 24-h urinary protein, which is inconvenient during follow-up. Objective To quickly predict the severity of CKD using more easily available demographic and blood biochemical features during follow-up, we developed and compared several predictive models using statistical, machine learning and neural network approaches. Methods The clinical and blood biochemical results from 551 patients with proteinuria were collected. Thirteen blood-derived tests and 5 demographic features were used as non-urinary clinical variables to predict the 24-h urinary protein outcome response. Nine predictive models were established and compared, including logistic regression, Elastic Net, lasso regression, ridge regression, support vector machine, random forest, XGBoost, neural network and k-nearest neighbor. The AU-ROC, sensitivity (recall), specificity, accuracy, log-loss and precision of each of the models were evaluated. The effect sizes of each variable were analysed and ranked. Results The linear models including Elastic Net, lasso regression, ridge regression and logistic regression showed the highest overall predictive power, with an average AUC and a precision above 0.87 and 0.8, respectively. Logistic regression ranked first, reaching an AUC of 0.873, with a sensitivity and specificity of 0.83 and 0.82, respectively. The model with the highest sensitivity was Elastic Net (0.85), while XGBoost showed the highest specificity (0.83). In the effect size analyses, we identified that ALB, Scr, TG, LDL and EGFR had important impacts on the predictability of the models, while other predictors such as CRP, HDL and SNA were less important. Conclusions Blood-derived tests could be applied as non-urinary predictors during outpatient follow-up. Features in routine blood tests, including ALB, Scr, TG, LDL and EGFR levels, showed predictive ability for CKD severity. The developed online tool can facilitate the prediction of proteinuria progress during follow-up in clinical practice. Electronic supplementary material The online version of this article (10.1186/s12967-019-1860-0) contains supplementary material, which is available to authorized users.

Comparison and development of machine learning tools for the prediction of chronic obstructive pulmonary disease in the Chinese population

“…We evaluated the predictive performance of these 6 models for COPD and found that the XGboost model presented the best AU-ROC and AU-PRC values in both the training and test sets in all features. The XGboost algorithm is a highly effective and widely used machine learning method that can build complex models and make accurate decisions when given adequate data [65]. We used the XGboost model to predict feature importance, and the results indicated that the AQCI was the most important factor, while SNPs were less important.…”

“…For example, the linear LR model [64] and SVM model have been widely adopted in many clinical applications, such as for CKD disease prediction [65]. The DT model [66] is based on a radial basis function neural network and support vector machine coupled with firefly algorithm techniques; the XGboost and MLP models have also been used in clinical research [65,67]. KNN was chosen due to its simplicity and ability to perform multiclass classification, and this algorithm could run with default parameters [68].…”

Chronic Kidney Disease Diagnosis using Decision Tree Algorithms

“…Their study obtained 85.3% accuracy to detect CKD. In 2019, Jing Xiao conducted a study to detect various stages of CKD [8]. This study used the logistic regression machine learning technique to train the model and used online tool for prediction.…”

“…Their comparative study revealed that Radial Basis Function provides the best accuracy rate with 85.3 percentage. Jing Xiao [8] established nine models and compared their performance to predict the CKD stages according to its severity. Predictive models include ridge regression, lasso regression, logistic regression, Elastic Net, XG Boost, neural network, k-nearest neighbor, random forest and support vector machine.…”

A Deep Learning Model for Estimation of Patients with Undiagnosed Diabetes

“…Previous studies have introduced predictive models for diseases such as diabetic retinopathy, skin cancer, lung disease, heart failure, chronic kidney disease, and so on using machine learning techniques [14][15][16][17][18][19][20]. These studies that use deep learning techniques to make major advances in solving problems have resisted the best attempts of the artificial intelligence community in many cases [21].…”