Hybrid positron emission tomography/computed tomography (PET/CT) and single-photon emission computed tomography/computed tomography (SPECT/CT) have evolved from predominantly qualitative techniques into quantitative imaging modalities capable of providing reproducible imaging biomarkers for diagnosis, treatment response assessment, radionuclide therapy planning and multicenter clinical research. The clinical reliability of quantitative imaging biomarkers depends on reducing biological and technical variability throughout the imaging workflow, including patient preparation, scanner calibration, acquisition protocols, image reconstruction and quantitative analysis. For the purposes of this review, standardization refers primarily to the implementation of validated and reproducible procedures within a defined imaging workflow or protocol, whereas harmonization aims to reduce scanner-, reconstruction-, calibration- and protocol-related variability so that quantitative biomarkers meet predefined comparability specifications across imaging systems and centers. This distinction provides a practical operational framework rather than an absolute boundary. The European Association of Nuclear Medicine (EANM) Research Ltd. (EARL) accreditation program represents the most widely adopted framework for PET/CT harmonization; formal standards for quantitative SPECT/CT are now emerging, including the EARL ¹⁷⁷Lu SPECT/CT accreditation program, while dedicated validation remains necessary for cadmium–zinc–telluride (CZT)-based imaging and radionuclide therapy dosimetry. This review summarizes the principles of quantitative PET/CT and SPECT/CT, the principal sources of quantitative variability, current standardization and harmonization strategies, and emerging challenges associated with digital PET, CZT technology, radiomics and artificial intelligence. Practical recommendations for implementation, quality assurance and structured quantitative reporting are also presented. Ultimately, quantitative imaging should be regarded as a standardized measurement system capable of providing accurate, reproducible and clinically interpretable imaging biomarkers. Standardization and harmonization are complementary processes that together provide the methodological foundation for precision nuclear medicine, multicenter clinical trials and future data-driven clinical practice. The future of quantitative molecular imaging will depend not only on technological innovation but also on maintaining comparable quantitative performance and clinically consistent interpretation across imaging platforms and clinical environments.
BACKGROUND: Deep-learning applications in nuclear medicine are expanding, but many departments lack dedicated AI engineering support, creating a gap between interest and practical implementation. The aim of this study was to evaluate the feasibility of a physician-led, incremental workflow for developing reproducible deep-learning prototypes for nuclear medicine imaging across increasing data complexity (2D planar, sequential planar “3D-like”, and true 3D DICOM) and to quantify prototype discrimination where reference standards were available. MATERIAL AND METHODS: Models were developed in Google Colab Pro using TensorFlow/Keras in a three-step framework. Step 1 used exported planar whole-body scintigraphy images (287 bone scintigraphy; 256 besilesomab scintigraphy) for radiopharmaceutical discrimination and an exploratory positive/negative subclassification without an external reference standard. Step 2 used ordered planar V/Q image stacks from 287 patients; reports provided reference-standard labels for acute pulmonary embolism (perfusion) and ventilation defects. Step 3 used true 3D PET DICOM volumes (46 [¹⁸F]FDG, 42 [¹⁸F]FDOPA, 42 [¹⁸F]FCH) with report-derived radiopharmaceutical labels. Datasets were split into 80% training and 20% validation subsets; performance was reported on the held-out validation subset unless otherwise stated. Primary outcome was feasibility (end-to-end pipeline completion and reproducible inference from an exported model); secondary outcomes included accuracy and AUC (plus sensitivity/specificity for Step 2). RESULTS: Feasibility criteria were met across all three steps, with trained models exported and reloaded for inference in a separate application context. Step 1 radiopharmaceutical discrimination was successful in a small illustrative evaluation set (n = 20 images), while positive/negative subclassification demonstrated reduced reliability consistent with label ambiguity (exploratory). In Step 2, the perfusion model achieved an accuracy of 0.85 and an AUC of 0.96, with a sensitivity of 96.6% and a specificity of 86.6%; the ventilation model achieved an accuracy of 0.82 and an AUC of 0.93, with a sensitivity of 88.4% and a specificity of 76.9%. In Step 3, the 3D ResNet34 achieved a validation accuracy of 1.00 and an AUC of 0.95 for radiopharmaceutical discrimination. CONCLUSIONS: This study demonstrates that deep-learning-based AI can be feasibly implemented even in small nuclear medicine departments. Although prototype in nature, it outlines a complete workflow from data extraction to model training and inference. Crucially, it highlights key limitations — such as reference-standard quality, task definition, preprocessing sensitivity, and the absence of an independent test set — that must be addressed to obtain a reliable and clinically meaningful in-house model.
