Meeting Banner
Abstract #3631

Tissue Expansion Maps (TEMs) Derived from Nonlinear Registration of Serial 3D MR Scans as an Imaging Biomarker for Detecting Brain Tumor Invasion and Quantifying Tumor Response to Therapy

Benjamin M. Ellingson1, 2, Timothy F. Cloughesy3, Robert J. Harris1, Davis C. Woodworth1, Kevin Leu1, Albert Lai3, Phioanh (Leia) Nghiemphu3, Whitney B. Pope1

1Dept. of Radiological Sciences, UCLA, Los Angeles, CA, United States; 2Biomedical Physics, UCLA, Los Angeles, CA, United States; 3Neurology, UCLA, Los Angeles, CA, United States

Expert radiologists and neuro-oncologists can identify subtle growth in the tumor through visual examination of the mass effect on serial MR scans; however, these evaluations are subjective and are not quantitative. The current study involves the development of a new class of imaging biomarkers that quantify parameters associated tissue deformation fields derived from serial nonlinear (elastic) registration of high-resolution post-contrast T1-weighted images in patients with glioblastoma. Tissue Expansion Maps (TEMs) that quantify the local tissue displacement velocity and local distortion vector field maps were can be used to detect new tumor invasion and understand patterns in tumor growth and infiltration.

Abstract PDF

Keywords

aggressive apparent audience bidirectional biomedical blood board brain capsule clinical common consists contain contrast corpus cost criteria cross currently deep deformation degree degrees depicting detecting determinant diffuse displacement distance distortion dynamic edge elastic elevated enhancement enhancing euclidean evaluated evaluation examining expansion explored extent extracted field fields fine flair form free freedom frontal function gauss gray grew growing growth hypothesized identified illustrates indicate individual infiltrating infiltration informed institutional internal invasion inversion lateral leading least lesion lobe local locus malignant many maps mass matrices matrix maximal medial median minimal months mutual near neurological neurosurgeons newly newton nodule nonlinear often optimizing partially participating particular patient patients physics poor pope portion poses post prognosis quantification quantified quantify quantifying radiological rapid received recurrent reflecting registered registration resection resolution response sciences serial serially signature signed similarly since slightly squares structures subsequent subtle suggesting susceptibility target therapy tissue tool treatments tumor typical uptake various vector visualization volumetric wise