DOI: 10.3390/diagnostics16132033 ISSN: 2075-4418

Technical Optimization Strategies for Amyloid PET Under Challenging Acquisition Conditions: A Comprehensive Narrative Review

Luca Camoni, Francesco Dondi, Agata Pietrzak, Roberto Rinaldi, Francesca Tomasoni, Michela Cossandi, Silvia Lucchini, Gian Luca Viganò, Luigi Spiazzi, Francesco Bertagna

Amyloid PET is increasingly used to confirm cerebral amyloid burden, but standard acquisition may be compromised by head motion, limited patient cooperation, reduced effective counts, premature scan termination, or non-repeatable imaging conditions. This comprehensive narrative review used a structured evidence-mapping approach in accordance with SANRA quality criteria. A structured literature search was performed in PubMed/MEDLINE, Scopus, and Web of Science up to 15 March 2026. Eligible studies included clinical, phantom, or hybrid studies addressing acquisition-time reduction, injected-activity reduction or low-count imaging, motion correction, or artificial intelligence-based image enhancement. Findings were synthesized narratively because of substantial heterogeneity in tracers, scanners, protocols, reconstruction methods, populations, comparators, and endpoints. Sixteen studies were included. Moderate reductions in acquisition time or effective counts generally preserved semiquantitative performance, whereas visual interpretation became more vulnerable under more aggressive reductions, borderline amyloid status, or reduced image quality. Artificial intelligence-based restoration improved image-quality metrics and supported interpretation of short- or low-count acquisitions, but evidence remained model-specific. Motion correction was supported by one amyloid-specific [18F]flutemetamol PET/CT study and should be interpreted as a potentially useful but under-replicated strategy. Current evidence supports cautious, tracer-, scanner-, reconstruction-, and task-specific optimization under challenging acquisition conditions rather than universal protocol reduction or direct generalization to motion-prone, poorly cooperative, or non-repeatable acquisition scenario. Reduced protocols, artificial intelligence-based restoration, and motion correction should remain locally validated supportive strategies, not substitutes for standard acquisition.

More from our Archive