Instead of the "one-dose-fits-all" treatment approach currently used for glioblastoma multiforme, presenter Pallavi Tiwari, PhD, and colleagues hoped to identify prognostic markers to provide personalized therapy for these patients.
Their hypothesis was that the heterogeneity in glioblastoma multiforme -- due to variations in enhancement, cellular density, necrosis, and fibrosis -- can be captured using computerized texture descriptors extracted from within different tumor compartments. The combination of these computerized texture features could then be used to predict short-term versus long-term survival.
The texture features across compartments were more prognostic of clinical survival than features from enhancing tumor and tumor volume alone, the group found.
Check out this Sunday presentation to get all the details.
