Machine Learning methods for Quantitative Radiomic Biomarkers.

Chintan Parmar,Patrick Grossmann,Johan Bussink,Philippe Lambin,Hugo J W L Aerts

doi:10.1038/srep13087

Abstract

Radiomics extracts and mines large number of medical imaging features quantifying tumor phenotypic characteristics. Highly accurate and reliable machine-learning approaches can drive the success of radiomic applications in clinical care. In this radiomic study, fourteen feature selection methods and twelve classification methods were examined in terms of their performance and stability for predicting overall survival. A total of 440 radiomic features were extracted from pre-treatment computed tomography (CT) images of 464 lung cancer patients. To ensure the unbiased evaluation of different machine-learning methods, publicly available implementations along with reported parameter configurations were used. Furthermore, we used two independent radiomic cohorts for training (n = 310 patients) and validation (n = 154 patients). We identified that Wilcoxon test based feature selection method WLCX (stability = 0.84 ± 0.05, AUC = 0.65 ± 0.02) and a classification method random forest RF (RSD = 3.52%, AUC = 0.66 ± 0.03) had highest prognostic performance with high stability against data perturbation. Our variability analysis indicated that the choice of classification method is the most dominant source of performance variation (34.21% of total variance). Identification of optimal machine-learning methods for radiomic applications is a crucial step towards stable and clinically relevant radiomic biomarkers, providing a non-invasive way of quantifying and monitoring tumor-phenotypic characteristics in clinical practice.

Highlights

Radiomics extracts and mines large number of medical imaging features quantifying tumor phenotypic characteristics
To investigate the machine-learning approaches for prognostic radiomic biomarkers, a total of 440 radiomic features were extracted from the segmented tumor regions of the pre-treatment computed tomography (CT) images of two independent NSCLC cohorts
Predictive performance of different feature selection and classification methods was assessed using the area under receiver operator characteristic curve (AUC)

Summary

Introduction

Radiomics extracts and mines large number of medical imaging features quantifying tumor phenotypic characteristics. Accurate and reliable machine-learning approaches can drive the success of radiomic applications in clinical care In this radiomic study, fourteen feature selection methods and twelve classification methods were examined in terms of their performance and stability for predicting overall survival. Few recent studies have investigated the effect of different feature selection and machine learning classification methods on radiomics based clinical predictions[19,20], but with limited sample sizes. These studies lacked independent validation of the results, which may restrict the generalizability of their conclusions

Objectives

Methods

Results

Conclusion