NIMG-52. CORRELATION OF BIDIMENSIONAL AND VOLUMETRIC TUMOR SIZING IN NEWLY DIAGNOSED AND RECURRENT GLIOBLASTOMA

• DOI: 10.1093/neuonc/noad179.0748
• Publication Date: 2023-11-11T23:24:29Z
• Authors: John Kim; Thomas Nelson; Gilbert Youssef; Rifaquat Rahman; Omar Arnaout; Wenya Linda Bi; Daniel Cagney; Yuh-Shin Chang; Matthew DeSalvo; Thomas Flood; Elizabeth Gerstner; Luis Nicolas Gonz...; Jeffrey Guenette; Albert Kim; Eudocia Lee; Jose Ricardo McFa...; Christopher Potter; David Reardon; Patrick Wen; Raymond Y Huang
• Keywords: 0301 basic medicine; 03 medical and health sciences; 3. Good health

Abstract BACKGROUND Glioblastoma (GBM) is the most common malignant primary nervous system tumor and remains incurable with a poor prognosis despite current therapy. The standard for monitoring response to treatment, as defined by Response Assessment in Neuro-Oncology (RANO) criteria, relies on bidimensional (2D) measurement product of T1-gadolinium enhancing disease. However, due variability characterization, there considerable interest volumetric segmentation assess burden, which could be implemented using machine learning models offering workflow automatization. METHODS In this study, we evaluate commercially available automated tool correlate its output manual assessment 2D burden experts. MRI examinations during systemic treatment two retrospective cohorts patients either newly diagnosed (nGBM) or recurrent GBM (rGBM) were assessed. diameter cm^2 was determined blinded readers segmentations cm^3 automatically generated. Spearman's correlation between volume measurements calculated each cohort. RESULTS 315 nGBM, 482 rGBM evaluated, comprising 3073 total brain MRIs nGBM cohort 2439 0.7463 (0.730, 0.762) 0.8598 (0.849, 0.870) CONCLUSION Automated tools have potential revolutionize radiology workflows increase sensitivity subtle changes imaging. differential performance confirms need further refinement customization such account tumor- patient-specific factors.