U-Net-based SUV calculation in FDG-PET imaging of mice brain for enhanced analysis

Abstract Positron emission tomography (PET) is a widely used imaging modality in nuclear medicine for variety of applications. Amongst the methods quantifying and interpretation PET images, standardized uptake value (SUV) widely-adopted semi-quantitative tool that supplements visual understanding with quantitative information. SUV both clinical preclinical practices to report status various normal organs tumors under investigation using imaging. While determination SUVs typically done manually, which can be tedious, artificial intelligence (AI) utilized enhance efficiency process. In this study, U-Net-based approach was employed semi-automated FDG-PET scans mice brains. First, U-Net model trained 50 images six perform automatic segmentation task. The then delineated brain mouse processed by short in-house code extract data calculate SUV. process also replicated manual way comparison purposes. results from method conventional at nine different time points revealed there were errors less than 4.5% eight out nine-time points. Although our model’s performance needs improvement, adapting well-trained AIbased determination, particularly studies, help reduce workload organ delineation minimize associated errors, facilitating determination.