Machine learning for grading prediction and survival analysis in high grade glioma.

in Scientific reports by Xiangzhi Li, Xueqi Huang, Yi Shen, Sihui Yu, Lin Zheng, Yunxiang Cai, Yang Yang, Renyuan Zhang, Lingying Zhu, Enyu Wang

TLDR

  • A magnetic resonance imaging (MRI)-based radiomics model was developed and validated for high-grade glioma (HGG) classification, achieving high performance with XGBoost and Stacking fusion models.
  • The study extracted radiomics features from T1-weighted imaging (T1WI) and compared seven classification methods, including logistic regression and tree-based methods.

Abstract

We developed and validated a magnetic resonance imaging (MRI)-based radiomics model for the classification of high-grade glioma (HGG) and determined the optimal machine learning (ML) approach. This retrospective analysis included 184 patients (59 grade III lesions and 125 grade IV lesions). Radiomics features were extracted from MRI with T1-weighted imaging (T1WI). The least absolute shrinkage and selection operator (LASSO) feature selection method and seven classification methods including logistic regression, XGBoost, Decision Tree, Random Forest (RF), Adaboost, Gradient Boosting Decision Tree, and Stacking fusion model were used to differentiate HGG. Performance was compared on AUC, sensitivity, accuracy, precision and specificity. In the non-fusion models, the best performance was achieved by using the XGBoost classifier, and using SMOTE to deal with the data imbalance to improve the performance of all the classifiers. The Stacking fusion model performed the best, with an AUC = 0.95 (sensitivity of 0.84; accuracy of 0.85; F1 score of 0.85). MRI-based quantitative radiomics features have good performance in identifying the classification of HGG. The XGBoost method outperforms the classifiers in the non-fusion model and the Stacking fusion model outperforms the non-fusion model.

Overview

  • The study aimed to develop and validate an MRI-based radiomics model for the classification of high-grade glioma (HGG) and determine the optimal machine learning (ML) approach.
  • The analysis included 184 patients with 59 grade III and 125 grade IV lesions, using T1-weighted imaging (T1WI) features.
  • The study compared the performance of seven classification methods, including logistic regression, XGBoost, Decision Tree, and others, with and without feature selection and balancing.

Comparative Analysis & Findings

  • The XGBoost classifier performed best in non-fusion models, achieving an AUC of 0.92 with SMOTE feature balancing.
  • The Stacking fusion model achieved the highest overall performance, with an AUC of 0.95, sensitivity of 0.84, accuracy of 0.85, and F1 score of 0.85.
  • The radiomics features extracted from MRI T1WI showed good performance in identifying HGG classification.

Implications and Future Directions

  • This study demonstrates the potential of MRI-based radiomics for identifying HGG, with implications for improved diagnosis and treatment outcomes.
  • Future studies can investigate the generalizability of this approach to other brain tumors and the potential role of radiomics in monitoring treatment response.
  • The Stacking fusion model's high performance suggests its potential as a valuable tool in clinical decision-making, but further validation is needed.
Read Full Article