Predicting survival in glioblastoma with multimodal neuroimaging and machine learning

Abstract

Purpose Glioblastoma (GBM) is the most common and aggressive malignant glioma, with an overall median survival of less than two years. The ability to predict survival before treatment in GBM patients would lead to improved disease management, clinical trial enrollment, and patient care. Methods GBM patients (N = 133, mean age 60.8 years, median survival 14.1 months, 57.9% male) were retrospectively recruited from the neurosurgery brain tumor service at Washington University Medical Center. All patients completed structural neuroimaging and resting state functional MRI (RS-fMRI) before surgery. Demographics, measures of cortical thickness (CT), and resting state functional network connectivity (FC) were used to train a deep neural network to classify patients based on survival (< 1y, 1-2y, >2y). Permutation feature importance identified the strongest predictors of survival based on the trained models. Results The models achieved a combined cross-validation and hold out accuracy of 90.6% in classifying survival (< 1y, 1-2y, >2y). The strongest demographic predictors were age at diagnosis and sex. The strongest CT predictors of survival included the superior temporal sulcus, parahippocampal gyrus, pericalcarine, pars triangularis, and middle temporal regions. The strongest FC features primarily involved dorsal and inferior somatomotor, visual, and cingulo-opercular networks. Conclusion We demonstrate that machine learning can accurately classify survival in GBM patients based on multimodal neuroimaging before any surgical or medical intervention. These results were achieved without information regarding presentation symptoms, treatments, postsurgical outcomes, or tumor genomic information. Our results suggest GBMs have a global effect on the brain’s structural and functional organization, which is predictive of survival.

Speeding up Glioblastoma Cancer Research: Highlighting the Zebrafish Xenograft Model

“…GBM is a diffuse glioma that originates from astrocytes cells, and it can arise in any area of the central nervous system (CNS). However, they occur more frequently in the cerebral hemispheres [12]. The GBM cells present a polymorphic morphology with rounded or spindle-shaped cells, or very small or very large cells, with marked nuclear atypia and intense mitotic activity.…”

“…The GBM cells present a polymorphic morphology with rounded or spindle-shaped cells, or very small or very large cells, with marked nuclear atypia and intense mitotic activity. Furthermore, an intense microvascular proliferation and large areas of necrosis are observed in the neoplasm area [12,13]. In the 2021 brain tumor classification, the World Health Organization (WHO), integrating histological features and molecular parameters, classified GBM as isocitrate dehydrogenase (IDH) wild-type and histone H3 wild-type diffuse glioma.…”

Radiogenomics-Based Risk Prediction of Glioblastoma Multiforme with Clinical Relevance

“…However, its predictive potential in PsP remains debatable [ 14 , 19 , 20 , 21 , 22 ]. Some machine learning and deep neural network methods that can predict the survival rate of GBM patients based on imaging data and clinical features have been developed [ 23 , 24 , 25 , 26 ]. However, there are no systematical and comprehensive approaches for the accurate stratification of PsP and TTP and the prediction of clinical outcomes of GBM patients receiving standard treatment.…”

Advancing precision prognostication in neuro-oncology: Machine learning models for data-driven personalized survival predictions in IDH-wildtype glioblastoma

“… 14 , 15 Second, ML facilitates integrating diverse variables spanning clinical, genomic, and imaging parameters, enabling a more comprehensive, personalized prognosis forecasting. 16 , 17 Finally, ML eliminates the need for rigid assumptions mandated by classical models, enabling the identification of nonlinear interactions and associations within high-dimensional data. 14 , 15 Collectively, these capabilities make ML a potentially more robust methodology compared to conventional statistics for generating individualized survival predictions in GBM patients.…”