Artificial Neural Network Accurately Predicts Hepatitis B Surface Antigen Seroclearance

Abstract

Background & AimsHepatitis B surface antigen (HBsAg) seroclearance and seroconversion are regarded as favorable outcomes of chronic hepatitis B (CHB). This study aimed to develop artificial neural networks (ANNs) that could accurately predict HBsAg seroclearance or seroconversion on the basis of available serum variables.MethodsData from 203 untreated, HBeAg-negative CHB patients with spontaneous HBsAg seroclearance (63 with HBsAg seroconversion), and 203 age- and sex-matched HBeAg-negative controls were analyzed. ANNs and logistic regression models (LRMs) were built and tested according to HBsAg seroclearance and seroconversion. Predictive accuracy was assessed with area under the receiver operating characteristic curve (AUROC).ResultsSerum quantitative HBsAg (qHBsAg) and HBV DNA levels, qHBsAg and HBV DNA reduction were related to HBsAg seroclearance (P<0.001) and were used for ANN/LRM-HBsAg seroclearance building, whereas, qHBsAg reduction was not associated with ANN-HBsAg seroconversion (P = 0.197) and LRM-HBsAg seroconversion was solely based on qHBsAg (P = 0.01). For HBsAg seroclearance, AUROCs of ANN were 0.96, 0.93 and 0.95 for the training, testing and genotype B subgroups respectively. They were significantly higher than those of LRM, qHBsAg and HBV DNA (all P<0.05). Although the performance of ANN-HBsAg seroconversion (AUROC 0.757) was inferior to that for HBsAg seroclearance, it tended to be better than those of LRM, qHBsAg and HBV DNA.ConclusionsANN identifies spontaneous HBsAg seroclearance in HBeAg-negative CHB patients with better accuracy, on the basis of easily available serum data. More useful predictors for HBsAg seroconversion are still needed to be explored in the future.

Bayesian network to predict hepatitis B surface antigen seroclearance in chronic hepatitis B patients

“…Previously, conventional regression and machine-learning analyses, based on single or a limited number of predictors, have shown inferior or comparable performance at predicting HBsAg seroclearance compared with our BN model. [8][9][10][11][12][13][14][15][16][17][18] However, with the exception of two studies that had a similar number of participants to those of the present study, 8,17 these previous studies had small sample sizes and limited inclusion of HBeAg seroconverters or HBeAg-negative CHB patients. 8,9,13,15,16 Furthermore, they did not take complex interactions into account, and the results of machine learning are often uninterpretable.…”

“…Predictive factors and models for HBsAg seroclearance have attracted much attention recently, with previous studies focusing on developing conventional statistical models, such as Cox regression analysis, [8][9][10] logistic regression analysis 11 and receiver operating characteristic (ROC) curve analysis [12][13][14][15][16] or machine-learning algorithms, 17,18 such as artificial neural network, extreme gradient boosting, random forest and decision tree to predict HBsAg seroclearance for CHB patients or subgroups such as hepatitis Be antigen (HBeAg)-seronegative CHB patients, using a single or a limited number of predictors. However, first, ROC analysis including a single predictor may miss valuable information.…”

Using Machine Learning Algorithms to Predict Hepatitis B Surface Antigen Seroclearance

“…A model using serum quantitative HBsAg (qHBsAg) and HBV DNA levels as proven clinical parameters to predict HBsAg seroclearance and seroconversion has been developed previously with artificial neural networks (ANNs), which is the only existing model on HBsAg seroclearance patients according to the best of our knowledge [21]. However, some limitations should be noted in this study including the limitation of small datasets, requirement of longitudinal follow-up data, and limited information considered except the currently proven predictor qHBsAg, an appropriate model with sufficient accuracy and generalizability for early predicting HBsAg seroconversion remains to be provided.…”

A model with combined viral and metabolic factors effectively predicts HBeAg status under long term entecavir therapy: a prospective cohort study

“…The cut-off values for the stratification of individual probability were decided by considering the natural history of CHB and recognized efficacy of ETV long-term treatment. According to prior studies, the annual incidence of spontaneous HBeAg seroconversion is about 2–15 %, and thus 20 % was chosen as the lower cut-off value to exclude spontaneous HBeAg seroconversion [ 37 ]. The probability of HBeAg seroconversion under ETV five years’ therapy is approximately 40 % or more [ 8 ], thus this value was selected as the upper cut-off line.…”