RBIO-01. CHARACTERIZING THE TEMPORAL EVOLUTION OF BRAIN ARCHITECTURE DURING RADIOTHERAPY FOR GLIOBLASTOMA USING NONLINEAR IMAGE REGISTRATION

Abstract BACKGROUND Previous serial MRI studies following glioblastoma (GBM) resection reveal dynamic changes in the tumor cavity and surrounding tissue. However, the time course of these dynamics is unclear, especially in the immediate postoperative period. METHODS GBM patients (n=8) received three MRIs: within 72 hours post-resection, immediately preceding RT (RTstart), and at the RT midpoint (RTmid). The radiation field (RF) was based on the resection cavity (RC) and surrounding T2-FLAIR hyperintensities at the 72 hour MRI. Nonlinear image registration was performed across the two time intervals, allowing comparison between the intended radiation targets and the actual irradiated tissue. Dice coefficients (DRF and DRC), the relative volume of the RC (RCvol), and the volume of unintended iatrogenic radiation (IRvol) were measured at each timepoint, and IRvol was fitted to an exponential decay. RESULTS DRF showed modest reduction, with mean±SD of 0.97±0.02 at both RTstart and RTmid. In contrast, the RC showed more change, with mean DRC 0.63±0.13 and 0.60±0.12, and mean RCvol of 0.74±0.27 and 0.72±0.25 at RTstart and RTmid, respectively. DRF, DRC, and RCvol showed greater change during the first interval than the second (p=3.5x10-4, 4.1x10-4, and 0.1, respectively). The mean IRvol was 16.7±12.6 at RTstart and 21.7±12.8 at RTmid, with a time constant of -0.031 (95% CI=[-0.017, -0.060]), indicating that 80% of all voxels that receive unintended iatrogenic radiation moved into the radiation field by postoperative day 57, and 95% will have moved in by day 125. CONCLUSIONS Nonlinear registration demonstrated significant postoperative architectural changes, with most occurring in the first two months after surgery near the RC. This potentially suggests a role for adaptive planning during RT, particularly for protocols that focus on the RC rather than T2-FLAIR hyperintensities. Future prospective studies are needed to assess the role of adaptive radiotherapy for GBM.

Authors

Institutions

Publication Details

Journal
Neuro-Oncology
Published
2024-11-01
DOI
https://doi.org/10.1093/neuonc/noae165.1067
Primary Topic
Cell Image Analysis Techniques
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

RBIO-01. CHARACTERIZING THE TEMPORAL EVOLUTION OF BRAIN ARCHITECTURE DURING RADIOTHERAPY FOR GLIOBLASTOMA USING NONLINEAR IMAGE REGISTRATION

Y. Xu, Jack Grinband, Tony Wang, Nicholas DeWalt
Neuro-Oncology
Cell Image Analysis Techniques
article

RBIO-01. CHARACTERIZING THE TEMPORAL EVOLUTION OF BRAIN ARCHITECTURE DURING RADIOTHERAPY FOR GLIOBLASTOMA USING NONLINEAR IMAGE REGISTRATION

Y. Xu, Jack Grinband, Tony Wang, Nicholas DeWalt
article en

Abstract

Abstract BACKGROUND Previous serial MRI studies following glioblastoma (GBM) resection reveal dynamic changes in the tumor cavity and surrounding tissue. However, the time course of these dynamics is unclear, especially in the immediate postoperative period. METHODS GBM patients (n=8) received three MRIs: within 72 hours post-resection, immediately preceding RT (RTstart), and at the RT midpoint (RTmid). The radiation field (RF) was based on the resection cavity (RC) and surrounding T2-FLAIR hyperintensities at the 72 hour MRI. Nonlinear image registration was performed across the two time intervals, allowing comparison between the intended radiation targets and the actual irradiated tissue. Dice coefficients (DRF and DRC), the relative volume of the RC (RCvol), and the volume of unintended iatrogenic radiation (IRvol) were measured at each timepoint, and IRvol was fitted to an exponential decay. RESULTS DRF showed modest reduction, with mean±SD of 0.97±0.02 at both RTstart and RTmid. In contrast, the RC showed more change, with mean DRC 0.63±0.13 and 0.60±0.12, and mean RCvol of 0.74±0.27 and 0.72±0.25 at RTstart and RTmid, respectively. DRF, DRC, and RCvol showed greater change during the first interval than the second (p=3.5x10-4, 4.1x10-4, and 0.1, respectively). The mean IRvol was 16.7±12.6 at RTstart and 21.7±12.8 at RTmid, with a time constant of -0.031 (95% CI=[-0.017, -0.060]), indicating that 80% of all voxels that receive unintended iatrogenic radiation moved into the radiation field by postoperative day 57, and 95% will have moved in by day 125. CONCLUSIONS Nonlinear registration demonstrated significant postoperative architectural changes, with most occurring in the first two months after surgery near the RC. This potentially suggests a role for adaptive planning during RT, particularly for protocols that focus on the RC rather than T2-FLAIR hyperintensities. Future prospective studies are needed to assess the role of adaptive radiotherapy for GBM.

Neuro-OncologyVol. 26(Supplement_8)
Columbia University Irving Medical Center (US), Columbia University (US)
Sustainable cities and communities
Openalex Percentile: Top 87%
Cell Image Analysis Techniques
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.