Radiation Physics

PV QA 3 - Poster Viewing Q&A 3

TU_7_3186 - Characterization of Tumor Response with FDG-PET images for Adaptive Treatment of Lung Cancer Patients

Tuesday, October 23
1:00 PM - 2:30 PM
Location: Innovation Hub, Exhibit Hall 3

Characterization of Tumor Response with FDG-PET images for Adaptive Treatment of Lung Cancer Patients
H. Zhong1, H. Sharifi2, F. M. Kong3, and I. J. Chetty1; 1Henry Ford Health System, Detroit, MI, 2Henry Ford Hospital, Detroit, MI, 3Indiana University, Bloomington, IN

Purpose/Objective(s): SUVpeak, SUVmax and total lesion glycolysis (TLG) are SUV-based metrics typically used for evaluation of treatment response for lung cancer patients. However, SUVmax is sensitive to image noise, SUVpeak depends on associated region of interest­ which may lead to uncertainties up to ±46%, and TLG cannot show region-specific SUV changes within a tumor. The purpose of this study was to develop a response-evaluation platform for measurement of radiation-induced, region-specific SUV changes potentially useful for adaptive treatment planning for lung cancer patients.

Materials/Methods: Suppose two sets of PET/CT images were acquired pre- and during-RT, respectively, with SUVs normalized to liver uptake and divided equally into 100 bins on pre-RT PET images. Let Vi denote the set of voxels associated with bin i, and Si denote the average SUV within Vi. Suppose the pre-RT CT images were registered to the during-RT CT first using an image-intensity-based B-Spline registration algorithm and then adjusted with a biomechanical model. The resultant displacement vector fields (DVFs) are then used to map the during-RT PET to the pre-RT PET using a mass-preserved projection method. The SUV differences between the pre-RT PET and the mapped during-RT PET are used to calculate the SUV change Ri in Vi. An accumulated response-SUV histogram at a point p, denoted by aRSH(p), 0≤p≤SUVmax, is then defined as the total of Ri divided by the total of Si, both summed with i satisfying Si≥p. Based on this definition, aRSH(p) corresponds to the relative change of SUVpeak with the peak volume defined by the union of Vi which satisfies Si≥p. If SUV=2.5 is selected as the tumor threshold, aRSH(2.5) is then equivalent to the relative change of TLG.

Results: The evaluation method was applied to five lung cancer patients. The mapped tumor volumes changed by 7.0±6.2% on average. The aRSH(p) calculated at p=2.5 was 63.7±15.2%, and 75.5±16.1% at p=SUVmax. The difference between the SUVmax of the pre-RT PET and that of the during-RT PET, denoted by ∆(SUVmax), was 46.0±34.3%, which is smaller than aRSH(SUVmax) because the two PET images do not necessarily have SUVmax values at the same point. The correlation coefficient between ∆(SUVmax) and aRSH(p) for the 5 patients was 0.152 at p=SUVmax, but increased to 0.838 as p decreased to 2.5. As p decreases, the accumulated Ri becomes less sensitive to registration errors, and consequently, aRSH(2.5) becomes more stable than aRSH(SUVmax).

Conclusion: A biomechanical model-based evaluation platform has been developed to evaluate radiation-induced SUV changes in individual bins. The derived aRSH curve may be beneficial toward: characterizing region-specific tumor response, analyzing patterns of tumor regression, and facilitating dose painting and response-based adaptive radiotherapy.

Author Disclosure: H. Zhong: None. H. Sharifi: None. F.(. Kong: None. I.J. Chetty: Research Grant; Varian Medical Systems, Inc, Philips Healthcare. Travel Expenses; Varian Medical Systems, Inc. ASTRO Nominating Committee member; ASTRO Nominating Committee.

Indrin Chetty, PhD, MS, FASTRO

Henry Ford Hospital

Presentation(s):

Send Email for Indrin Chetty


Assets

TU_7_3186 - Characterization of Tumor Response with FDG-PET images for Adaptive Treatment of Lung Cancer Patients



Attendees who have favorited this

Please enter your access key

The asset you are trying to access is locked. Please enter your access key to unlock.

Send Email for Characterization of Tumor Response with FDG-PET images for Adaptive Treatment of Lung Cancer Patients