Correlations of Complete Blood Count with Alanine and Aspartate Transaminase in Chinese Subjects and Prediction Based on Back-Propagation Artificial Neural Network (BP-ANN)

Abstract

BackgroundThe complete blood count (CBC) is the most common examination used to monitor overall health in clinical practice. Whether there is a relationship between CBC indexes and alanine transaminase (ALT) and aspartate aminotransferase (AST) has been unclear.Material/MethodsIn this study, 572 normal-weight and 346 overweight Chinese subjects were recruited. The relationship between CBC indexes with ALT and AST were analyzed by Pearson and Spearman correlations according to their sex, then we conducted colinearity diagnostics and multiple linear regression (MLR) analysis. A prediction model was developed by a back-propagation artificial neural network (BP-ANN).ResultsALT was related to 4 CBC indexes in the male normal-weight group and 3 CBC indexes in the female group. In the overweight group, ALT had a similar relationship with the normal group, but there was only 1 index related with AST in the normal-weight group and male overweight groups. The ALT regression models were developed in normal-weight and overweight people, which had better correlation coefficient (R>0.3). After training 1000 epochs, the BP-ANN models of ALT achieved higher correlations than MLR models in normal-weight and overweight people.ConclusionsALT is a more suitable index than AST for developing a regression model. ALT can be predicted by CBC indexes in normal-weight and overweight individuals based on a BP-ANN model, which was better than MLR analysis.

A prediction model of sleep disturbances among female nurses by using the BP‐ANN

“…The BP‐ANN does not require an explicit understanding of the mechanisms underlying the processes it examines but instead uses data sets (input, hidden and output data) pertaining throughout the mapping process. The BP‐ANN has been widely used in the medical field to diagnose diseases (Shichijo et al, ), quantify the synergism of drugs (Pivetta et al, ), predict the outcome of treatments (Qaderi et al, ) and explore the associations among indexes of assessed variables (Wang, Wang, & Liu, ; Yu et al, ). In contrast to binary logistic regression analysis, the process can minimize error for nonlinear functions of high complexity and allows for self‐tuning adaptive control (Buscema, ).…”

Comparison of Diagnosis Accuracy between a Backpropagation Artificial Neural Network Model and Linear Regression in Digestive Disease Patients: an Empirical Research

“…The linear regression for principal component analysis fitted the straight line which crosses the hidden layer in the neural network, and the next process was to generalize this straight line to a curve. Based on the principal component analysis, the BP-ANN model could fix nonlinear principal component data and the algorithm was backpropagation for mean square error (MSE) and composed of a gradient descent method which was widely used in numerical minimization of a preestablished cost function [ 38 , 46 ]. According to the gradient trends, the BP model could update parameters between hidden layers and the input layer [ 47 ].…”

“…The BP algorithm was based on error gradient descent ( Figure 1 ), which was aimed at finding the minimum error by adjusting weights of connections between neurons in the direction of lowest error [ 37 ]. The error was estimated from the output variable and backcalculated to converge to the optimum solutions [ 38 ].…”

Neural Network-Based Study about Correlation Model between TCM Constitution and Physical Examination Indexes Based on 950 Physical Examinees

“…The learning rate was ç . The excitation function was g ( x ) [ 46 ]. This study chose Sigmoid function (S-function) as the excitation function to establish the correlation model and to realize conversion of variable data and weight.…”