Clinical evaluation of tissue-dependent and spatially-variant positron range correction for Gallium-68 PET imaging.

Abstract

Positron range correction (PRC) can mitigate the effect of the larger positron range on the image quality of Gallium-68 (⁶⁸Ga) PET-imaging. The aim of this study is to evaluate the improvement in ⁶⁸Ga-PET image quality by applying a tissue-dependent and spatially-variant PRC (TDSV PRC) for ⁶⁸Ga in a clinical setting. A TDSV PRC technique was developed employing CT-driven segmentation masks of different tissue types (soft tissue, bone, lung) and the corresponding tissue-specific positron range kernels. OSEM reconstructions were performed using the proposed TDSV PRC, a tissue-independent PRC, and without any PRC (non-PRC). For lesions identified in [⁶⁸Ga]Ga-DOTATOC or [⁶⁸Ga]Ga-PSMA PET/CT data from 20 patients, the maximum standardized uptake value (SUV_max) and contrast-to-noise ratio (CNR) of this technique was compared to tissue-independent PRC and non-PRC images. A total of 93 lesions were analyzed (48 soft tissue, 35 bone, 10 lung lesions). For soft tissue lesions, TDSV and tissue-independent PRC showed similar increases in SUV_max (13.7%, p < 0.001 vs. 13.6%, p < 0.001) and CNR (11.0%, p < 0.001 vs. 11.1%, p < 0.001) compared to non-PRC. For bone lesions, tissue-independent PRC showed slightly higher not statistically significant increases than TDSV PRC in SUV_max (18.6%, p < 0.001 vs. 17.4%, p < 0.001) and CNR (14.6%, p < 0.001 vs. 13.8%, p < 0.001). In lung lesions, TDSV PRC increased SUV_max and CNR compared to non-PRC (SUV_max: 57.9%, p = 0.012; CNR: 43.9%, p = 0.012) and tissue-independent PRC (SUV_max: 46.0%, p = 0.012; CNR: 32.5%, p = 0.012). TDSV PRC for ⁶⁸Ga PET/CT demonstrated to be feasible in clinical patient data, showing the greatest benefits for lung lesions.