1. Multi-Parametric MRI and Texture Analysis to Visualize Spatial Histologic Heterogeneity and Tumor Extent in Glioblastoma
- Author
-
Stephen J. Price, Joseph C. Loftus, Robert B. Jenkins, Shuluo Ning, Brian P. O'Neill, Nhan L. Tran, Leslie C. Baxter, Jing Li, Nathan Gaw, William F. Elmquist, Fei Gao, Christine M. Zwart, Kristin R. Swanson, Jann N. Sarkaria, J. Ross Mitchell, John P. Karis, Jonathan D. Plasencia, David H. Frakes, Kris A. Smith, Amylou C. Dueck, Jennifer M. Eschbacher, Teresa Wu, Leland S. Hu, Sara Ranjbar, and Peter Nakaji
- Subjects
Image-Guided Biopsy ,Pathology ,medicine.medical_specialty ,lcsh:Medicine ,Contrast Media ,Biology ,Texture (geology) ,Machine Learning ,Parenchyma ,Biopsy ,Image Interpretation, Computer-Assisted ,medicine ,Humans ,lcsh:Science ,Multidisciplinary ,Multi parametric ,medicine.diagnostic_test ,Extramural ,lcsh:R ,Magnetic resonance imaging ,medicine.disease ,Magnetic Resonance Imaging ,Radiography ,Diffusion Tensor Imaging ,lcsh:Q ,Glioblastoma ,Algorithms ,Diffusion MRI ,Research Article - Abstract
Background Genetic profiling represents the future of neuro-oncology but suffers from inadequate biopsies in heterogeneous tumors like Glioblastoma (GBM). Contrast-enhanced MRI (CE-MRI) targets enhancing core (ENH) but yields adequate tumor in only ~60% of cases. Further, CE-MRI poorly localizes infiltrative tumor within surrounding non-enhancing parenchyma, or brain-around-tumor (BAT), despite the importance of characterizing this tumor segment, which universally recurs. In this study, we use multiple texture analysis and machine learning (ML) algorithms to analyze multi-parametric MRI, and produce new images indicating tumor-rich targets in GBM. Methods We recruited primary GBM patients undergoing image-guided biopsies and acquired pre-operative MRI: CE-MRI, Dynamic-Susceptibility-weighted-Contrast-enhanced-MRI, and Diffusion Tensor Imaging. Following image coregistration and region of interest placement at biopsy locations, we compared MRI metrics and regional texture with histologic diagnoses of high- vs low-tumor content (≥80% vs
- Published
- 2015