SU‐GG‐T‐491: Identifying Patients Who Will Potentially Benefit from Functional‐Guided Lung Avoidance in Thoracic Radiotherapy
Purpose: Single photon emission computed tomography (SPECT) may be used to guide thoracic radiation therapy planning by reducing the dose to highly‐perfused (i.e.functional) regions of lung. We herein investigate whether it is possible to a priori predict the extent of functional lung sparing possible with SPECT guidance, which can then be used to select patients who would most benefit from functional planning. Method and Materials: IMRT plans were generated with and without SPECT‐guidance for 6 patients. The dose‐perfusion (function) histograms were compared to quantify the degree of sparing of perfused lung afforded by inclusion of SPECT, i.e., the percent of perfusion at ≥ 20 Gy (F20) and 30 Gy (F30) were compared. In order to correlate the spatial distribution of SPECT intensity to the degree of dose sparing, three different SPECT‐descriptor metrics were considered as possible candidates: (a) second‐order moment invariants (J1, J2, J3), which are shape descriptors of the SPECT intensity; (b) overlap function histogram (OFH), which measures the perfusion within increasing concentric expansions of the PTV; (c) the percent of perfused lung within the beam outlines (FBEV). Results: The mean reductions in %F20 and %F30 from SPECT‐guidance were 11.20 (1.44–18.69) and 12.46 (−1.66–21.68), respectively. The correlation coefficients of %F20 reduction vs. J1, J2, J3, OFH and FBEV were −0.9272 (p=0.008), −0.8924 (p=0.055), −0.5910 (p=0.22), 0.5866 (p=0.22) and −0.1412 (p=0.79), respectively. The corresponding correlations for %F30 reduction were −0.9134 (p=0.011), −0.7332 (p=0.097), −0.5194 (p=0.29), 0.7262 (p=0.10) and 0.01371 (p=0.98), respectively. The strongest correlated metric, J1, may be used to predict %F30 and %F30 reductions as −0.1431*J1+35.147 (p=0.008) and −0.1719*J1+41.24 (p=0.011), respectively. Conclusion: J1, which sums the product of the SPECT intensity and square of the distance from the isocenter, can accurately predict those patients who will benefit from SPECT‐guided IMRT. © 2010, American Association of Physicists in Medicine. All rights reserved.
Seger, J; Marks, L; Das, S
Volume / Issue
Start / End Page
International Standard Serial Number (ISSN)
Digital Object Identifier (DOI)