Differentiating Radiation Necrosis and Metastatic Progression in Brain Tumors Using Radiomics and Machine Learning

TL;DR

This study develops an automated machine learning-based radiomics approach to accurately differentiate radiation necrosis from metastatic progression in brain tumors using MRI, potentially reducing the need for invasive biopsies.

Contribution

The paper introduces a novel radiomics and machine learning pipeline that achieves high accuracy in distinguishing RN from metastasis, improving diagnostic precision over conventional methods.

Findings

Random forest classifier achieved AUC of 0.910.

Gradient filter provided the best performance among tested features.

The approach can accurately differentiate RN from metastasis non-invasively.

Abstract

Objectives: Distinguishing between radiation necrosis(RN) and metastatic progression is extremely challenging due to their similarity in conventional imaging. This is crucial from a therapeutic point of view as this determines the outcome of the treatment. This study aims to establish an automated technique to differentiate RN from brain metastasis progression using radiomics with machine learning. Methods: 86 patients with brain metastasis after they underwent stereotactic radiosurgery as primary treatment were selected. Discrete wavelets transform, Laplacian-of-Gaussian, Gradient, and Square were applied to magnetic resonance post-contrast T1-weighted images to extract radiomics features. After feature selection, dataset was randomly split into train/test (80%/20%) datasets. Random forest classification(RFC), logistic regression, and support vector classification(SVC) were trained and subsequently validated using test set. The classification performance was measured by area under the curve(AUC) value of receiver operating characteristic curve, accuracy, sensitivity, and specificity. Results: The best performance was achieved using RFC with a Gradient filter (AUC=0.910, std=0.047), (accuracy 0.8, std=0.071), (sensitivity=0.796 std=0.055), (specificity =0.922, std=0.059). For SVC the best result obtains using wavelet_HHH with a high AUC of 0.890 with std=0.89, accuracy of 0.777 with std=0.062, sensitivity=0.701, std=0.084, and specificity=0.85 with std=0.112. Logistic regression using wavelet_HHH provides a poor result with AUC=0.882 & std=0.051, accuracy of 0.753 & std=0.08, sensitivity=0.717 & std=0.208, and specificity=0.816 with std=0.123. Conclusion: This type of machine-learning approach can help accurately distinguish RN from recurrence in magnetic resonance imaging, without the need for biopsy. This has the potential to improve the therapeutic outcome.