Machine Learning and Feature Selection Methods for EGFR Mutation Status Prediction in Lung Cancer

Joana Morgado,Beatriz Flor De Lima,Eduardo Negrão,Venceslau Hespanhol,Miguel Correia Da Silva,Cláudia Freitas,António Cunha,José Luis Costa,Isabel Ramos,António J Madureira,Hélder P Oliveira,Tania Pereira,Francisco Silva

doi:10.3390/app11073273

Abstract

The evolution of personalized medicine has changed the therapeutic strategy from classical chemotherapy and radiotherapy to a genetic modification targeted therapy, and although biopsy is the traditional method to genetically characterize lung cancer tumor, it is an invasive and painful procedure for the patient. Nodule image features extracted from computed tomography (CT) scans have been used to create machine learning models that predict gene mutation status in a noninvasive, fast, and easy-to-use manner. However, recent studies have shown that radiomic features extracted from an extended region of interest (ROI) beyond the tumor, might be more relevant to predict the mutation status in lung cancer, and consequently may be used to significantly decrease the mortality rate of patients battling this condition. In this work, we investigated the relation between image phenotypes and the mutation status of Epidermal Growth Factor Receptor (EGFR), the most frequently mutated gene in lung cancer with several approved targeted-therapies, using radiomic features extracted from the lung containing the nodule. A variety of linear, nonlinear, and ensemble predictive classification models, along with several feature selection methods, were used to classify the binary outcome of wild-type or mutant EGFR mutation status. The results show that a comprehensive approach using a ROI that included the lung with nodule can capture relevant information and successfully predict the EGFR mutation status with increased performance compared to local nodule analyses. Linear Support Vector Machine, Elastic Net, and Logistic Regression, combined with the Principal Component Analysis feature selection method implemented with 70% of variance in the feature set, were the best-performing classifiers, reaching Area Under the Curve (AUC) values ranging from 0.725 to 0.737. This approach that exploits a holistic analysis indicates that information from more extensive regions of the lung containing the nodule allows a more complete lung cancer characterization and should be considered in future radiogenomic studies.

Highlights

Lung cancer is the foremost determinant cancer death amongst men and women, killing a vaster number of people than colon, breast, and prostate cancers combined [1]
The results of this study suggest that the linear learning models—Support Vector Machine (SVM) with linear kernel, Elastic Net, and Logistic Regression—perform well with quantitative imaging features as their predictors, whereas the SVM classifier based on the radial basis function (RBF) kernel performs poorly
Predicting Epidermal Growth Factor Receptor (EGFR) mutation status by computed tomography (CT) imaging can improve the determination of the most appropriate treatment for each lung cancer patient and is a less invasive alternative compared to the traditional biopsy

Summary

Introduction

Lung cancer is the foremost determinant cancer death amongst men and women, killing a vaster number of people than colon, breast, and prostate cancers combined [1]. This is linked to the fact that it is often diagnosed in an advanced stage, with 5% or less chance of a 5-year survival [2]. The cell surface receptor EGFR is responsible for cell growth and survival and its mutations promote EGFR permanent activation, which contributes to uncontrolled cell division [5,6] This genomic biomarker with clinically approved therapies is considered a strong prognostic indicator in lung cancer, rising opportunities to explore treatment strategies that rely on the individual’s genetic profile [7].

Methods

Results

Discussion

Conclusion