Machine Learning and Radiomic Features to Predict Overall Survival Time for Glioblastoma Patients

Abstract

Glioblastoma is an aggressive brain tumor with a low survival rate. Understanding tumor behavior by predicting prognosis outcomes is a crucial factor in deciding a proper treatment plan. In this paper, an automatic overall survival time prediction system (OST) for glioblastoma patients is developed on the basis of radiomic features and machine learning (ML). This system is designed to predict prognosis outcomes by classifying a glioblastoma patient into one of three survival groups: short-term, mid-term, and long-term. To develop the prediction system, a medical dataset based on imaging information from magnetic resonance imaging (MRI) and non-imaging information is used. A novel radiomic feature extraction method is proposed and developed on the basis of volumetric and location information of brain tumor subregions extracted from MRI scans. This method is based on calculating the volumetric features from two brain sub-volumes obtained from the whole brain volume in MRI images using brain sectional planes (sagittal, coronal, and horizontal). Many experiments are conducted on the basis of various ML methods and combinations of feature extraction methods to develop the best OST system. In addition, the feature fusions of both radiomic and non-imaging features are examined to improve the accuracy of the prediction system. The best performance was achieved by the neural network and feature fusions.

Predicting Survival in Patients with Brain Tumors: Current State-of-the-Art of AI Methods Applied to MRI

“…Various authors focused on the combined use of ML and radiomic techniques for OS prediction in brain tumors [ 92 , 93 ]. Chato et al [ 92 ] focused on GBM, while Grist et al [ 93 ] focused on paediatric brain tumors.…”

“…Various authors focused on the combined use of ML and radiomic techniques for OS prediction in brain tumors [ 92 , 93 ]. Chato et al [ 92 ] focused on GBM, while Grist et al [ 93 ] focused on paediatric brain tumors. The latter combined multi-site MRI with ML methods to predict survival in paediatric brain tumors, with the aim of stratifying patients to low and high risk cohorts.…”

Risk Stratification and Overall Survival Prediction in Advanced Gastric Cancer Patients Based on Whole‐Volume MRI Radiomics

“…Thus, we assumed that RS could reflect a clinical potential in establishing a connection with MRI images and survival. 33,34 Some articles have indicated that pT and pN are very critical prognostic factors, which are instructive for OS in univariate Cox-regression analysis but not obvious in multivariate Cox-regression analysis. [35][36][37] A possible explanation for this observation was that the patient had underwent gastrectomy to remove the risk factors of the primary tumor, so the effect on the later survival is not obvious.…”

Prediction of Survival of Glioblastoma Patients Using Local Spatial Relationships and Global Structure Awareness in FLAIR MRI Brain Images

“…In the first approach, radiomics features are used for the prediction of survival. Several researchers [14,15,16,17,18,19,20] have used the BraTS dataset for training and evaluation. Deep features combined with radiomic (intensity, texture, wavelet, shape) and clinical information, are initially used, and then those features are input into a random forest regressor [14] to predict overall survival in days.…”

“…Another study used 4524 radiomics features and clinical data to train a random forest regressor [16]. Location features were investigated in [17] using a fusion of radiomics, location, and clinical features. A multilayer perceptron (MLP) network was then used for survival prediction using 13 shape features and clinical information [18].…”