Machine Learning Analytics of Resting-State Functional Connectivity Predicts Survival Outcomes of Glioblastoma Multiforme Patients

Abstract

Glioblastoma multiforme (GBM) is the most frequently occurring brain malignancy. Due to its poor prognosis with currently available treatments, there is a pressing need for easily accessible, non-invasive techniques to help inform pre-treatment planning, patient counseling, and improve outcomes. In this study we determined the feasibility of resting-state functional connectivity (rsFC) to classify GBM patients into short-term and long-term survival groups with respect to reported median survival (14.6 months). We used a support vector machine with rsFC between regions of interest as predictive features. We employed a novel hybrid feature selection method whereby features were first filtered using correlations between rsFC and OS, and then using the established method of recursive feature elimination (RFE) to select the optimal feature subset. Leave-one-subject-out cross-validation evaluated the performance of models. Classification between short- and long-term survival accuracy was 71.9%. Sensitivity and specificity were 77.1 and 65.5%, respectively. The area under the receiver operating characteristic curve was 0.752 (95% CI, 0.62–0.88). These findings suggest that highly specific features of rsFC may predict GBM survival. Taken together, the findings of this study support that resting-state fMRI and machine learning analytics could enable a radiomic biomarker for GBM, augmenting care and planning for individual patients.

Predicting overall survival in diffuse glioma from the presurgical connectome

“…No prior studies have evaluated T1w MRI -based connectome models of glioma survival. Our T1w MRI connectome model was superior to previously reported fMRI and DTI based connectome models in predicting glioma survival, which have demonstrated AUCs or accuracies of 0.75–0.81 11 , 12 , 46 . T1w MRI has distinct advantages compared to DTI in terms of reliability in constructing brain networks and standardizing pulse sequences across scanner types 47 , 48 .…”

Tumour-infiltrated cortex participates in large-scale cognitive circuits

“…Recent studies have also shown that intratumoural functional connectivity is prognostic of better survival and cognitive outcomes after surgery (21,38). The relationship between intratumoural functional connectivity and relevant clinical variables (e.g., overall survival, eloquence of tissue, and cognitive outcome) suggests that rs-fMRI could be a useful, noninvasive biomarker to help guide clinical decision making (21,38,39). Future research should help delineate the biological significance of intratumoural functional connectivity to inform the interpretability of BOLD signal in neoplastic tissue.…”

Brain network mapping and glioma pathophysiology

“…Indeed, some studies have found that measures of functional connectivity are predictive of clinical outcomes for glioma patients, independent of other known prognostic factors. 8 , 9 , 197 A related MEG study found that elevated oscillatory activity within diffuse gliomas was associated with poorer progression free survival outcomes, even after adjusting for potential confounding variables. 10 The molecular basis for these prognostic imaging markers is currently incompletely understood.…”