Artificial intelligence & nuclear medicine: an emerging partnership

This letter addresses a recently published article evaluating the performance of 3D U-Net–based deep learning models for automated lesion segmentation in PET/CT imaging. The study represents a significant advancement in the integration of artificial intelligence (AI) into Nuclear Medicine. By comparing volumetric, MIP-based, and hybrid segmentation approaches using [¹⁸F]FDG and [⁶⁸Ga]Ga-PSMA radiotracers, the authors demonstrate that hybrid models can enhance lesion detection and contouring accuracy. These findings underscore the potential of AI-based segmentation to improve consistency and reduce the manual workload in clinical PET/CT interpretation. We consider this work a pivotal step toward the clinical adoption of AI tools, offering tangible benefits for routine practice and radiomic analysis, while preserving the essential supervisory role of the Nuclear Medicine Physician.