Comparison of Logistic Regression and Artificial Neural Network Models in Breast Cancer Risk Estimation

Abstract

Computer models in medical diagnosis are being developed to help physicians differentiate between healthy patients and patients with disease. These models can aid in successful decision making by allowing calculation of disease likelihood on the basis of known patient characteristics and clinical test results. Two of the most frequently used computer models in clinical risk estimation are logistic regression and an artificial neural network. A study was conducted to review and compare these two models, elucidate the advantages and disadvantages of each, and provide criteria for model selection. The two models were used for estimation of breast cancer risk on the basis of mammographic descriptors and demographic risk factors. Although they demonstrated similar performance, the two models have unique characteristics-strengths as well as limitations-that must be considered and may prove complementary in contributing to improved clinical decision making.

A Comparison of Logistic Regression Analysis and an Artificial Neural Network Using the BI-RADS Lexicon for Ultrasonography in Conjunction with Introbserver Variability

“…LR and the ANN are the most frequently used computer models in clinical risk estimation [19]. The advantage in use of the ANN is the capacity to model complex non-linear relationships between independent and predictor variables.…”

“…If Y is denoted as an indicator of cancer, the probability of cancer, given x i , is as follows [19,26]:…”

“…Logistic regression (LR) analysis and an artificial neural network (ANN) have been applied for tumor characterization [18,19]. The LR model is superior for use in examination of possible causal relationships between independent and dependent variables and in understanding the effects of predictors on outcome variables [20].…”

Decision support systems for adoption in dental clinics: A survey

“…Though data learning and training in the hidden layer is not transparent to the user, ANN is simple to implement as it requires minimal statistical training. The logistic regression method utilises a simpler linear model, and unlike ANN which can handle arbitrary relationships between input and output variables, it can only be used if such relationships can be explicitly identified (Ayer et al, 2010). Thus, the logistic regression method is not as robust as ANN.…”

Feasibility Test of Quantitative Assessment of Breast Density Based on Dielectric Impedance Spectroscopy

“…For this purpose, a large number of breast cancer risk factors have been identified and investigated in the breast cancer screening field. In the cancer epidemiology field, a number of breast cancer risk assessment models have been identified, which rely primarily on the combination of several well-known risk factors, such as age, family history, information on special genotypes and breast density [4][5][6]. With the exception of age and carrying specific gene mutations that only apply to a very small fraction of the population, many studies have shown that the breast density is the strongest breast cancer risk indicator [7,8].…”