Artificial intelligence in personalized internal dosimetry for radiopharmaceutical therapy: a narrative mini-review

Personalized internal radiation dosimetry supports individualized radiopharmaceutical therapy but remains difficult to implement because workflows require repeated quantitative imaging, segmentation, registration, kinetic modeling, and absorbed-dose calculation. This mini-review summarizes current evidence on artificial intelligence across the personalized dosimetry workflow and considers clinical readiness. A narrative PubMed search was completed in early 2026 and covered peer-reviewed studies published from 2015 to early 2026. Studies addressing artificial intelligence, machine learning, or deep learning in quantitative image reconstruction, segmentation, registration, time-activity curve estimation, absorbed-dose calculation, or clinical decision support were considered. Studies reporting quantitative performance or validation results were prioritized. Deep learning reduced yttrium-90 single-photon emission computed tomography reconstruction error from 149.5% ± 31.2% to 9.2% ± 1.7%; lesion dose-rate error was 8.6% ± 5.7% in virtual-patient testing. Learning-based registration reduced computation time from about 50 s to under 6 s. Single-time-point approaches can simplify imaging, and recent machine learning work has explored prediction of organ effective half-life. Automated segmentation methods show improving reproducibility, whereas evidence for artificial intelligence-based clinical decision support remains mainly retrospective and single-center. Artificial intelligence applications in reconstruction and segmentation appear closer to clinical readiness, while decision-support applications still require prospective, multicenter validation.