TY - JOUR T1 - Compensation for partial volume effect in PET for treatment response assessment of stereotactic body radiotherapy JF - Journal of Nuclear Medicine JO - J Nucl Med SP - 2080 LP - 2080 VL - 52 IS - supplement 1 AU - Tinsu Pan AU - Xu Zhang AU - Adam Riegel AU - Joe Chang AU - Hubert Chuang Y1 - 2011/05/01 UR - http://jnm.snmjournals.org/content/52/supplement_1/2080.abstract N2 - 2080 Objectives PET is an effective tool for treatment response assessment. Partial volume (PV) causes underestimation of SUV for the lesions (< 3cm) in stereotactic body radiotherapy (SBRT). PV can be corrected by the recovery coefficient (RC) of the imaging system. However, there is no commercial software for this compensation. We developed a clinical utility to compensate for PV in PET, and assessed the corrections on 193 biopsy proven lung nodules for SBRT. The hypothesis is that SUV should not be a function of tumor size after PV correction. Methods We derived a RC as function of tumor size on the GE PET/CT scanner with a NEMA IEC phantom. The 6 spheres (diameters of 10 to 37 mm) and the background activity were filled to 6 signal-to-background ratios (SBR) of 5, 10, 15, 20, 30 and 50 to simulate the range of SBR in the patient data. The phantom at each SBR was scanned three times and 6 min per scan to improve statistics and to provide the calibration data for modeling the RC as a function of tumor size. We applied the RC to correct for PV on 193 biopsy proven lung lesions: 37 lesions < 1 cm, 112 lesions < 2 cm, and 44 lesions < 3 cm for SBRT. Tumor size measurements were made in the largest dimension of each tumor. Results Before correction for PV, the slope of the liner fit of SUV and the tumor size was 0.42, indicating an increase of SUV with an increase of tumor size. After correction for PV, the slope became 0.24, suggesting a decrease of correlation between SUV and tumor size, i.e., SUV is less dependent on tumor size. Two limitations in this study were not counting for tumor motion and not counting for different disease types of lung tumor. Conclusions We have developed a clinical utility to compensate for PV for treatment response assessment in SBRT. In the study of 193 lesions of less than 3 cm, we found the correlation between SUV and tumor size decreased after correction for PV, suggesting a less dependence of SUV with tumor size after correction for PV ER -