| Abstract: |
Positron emission tomography (PET) and computed tomography (CT) are widely used imaging modalities in radiation oncology. PET imaging has a high contrast but often leads to blurry tumor edges due to its limited spatial resolution, while CT imaging has a high resolution but a low contrast between a tumor and its surrounding normal soft tissues. Tumor segmentation from either a single PET or CT image is difficult. It is known that co-segmentation methods utilizing the complementary information between PET and CT can improve the segmentation accuracy. This complementary information can be either consistent or inconsistent in the image level. How to correctly localize tumor edges with the inconsistent information is one major challenge for co-segmentation. This paper proposes a novel variational method for tumor co-segmentation in PET/CT, with a fusion strategy specifically designed to handle the information inconsistency between PET and CT in an adaptive way-the method can automatically decide which modality should be more trustful when PET and CT disagree to each other for localizing the tumor boundary. A PET restoration process is further integrated into the co-segmentation process, which further eliminates the uncertainty for tumor segmentation introduced by tumor edge blur in PET. The performance of the proposed method is validated on 50 nonsmall cell lung cancer patients. It achieves an average dice similarity indexes (DSIs) of 0.85 +/- 0.06, volume error (VE) of 0.09 +/- 0.08, and classification error (CE) of 0.31 +/- 0.13. The proposed method has a high accuracy for PET/CT tumor co-segmentation and PET restoration. |
