Reston, VA (November 3, 2026)—In routine clinical practice, local radiologists and nuclear medicine physicians can reliably interpret amyloid PET scans to aid in the diagnosis and management of Alzheimer's disease, showing good agreement with interpretations by amyloid PET specialists. According to a new study published in The Journal of Nuclear Medicine, local interpretations of nearly 1,500 scans showed excellent agreement with expert panel assessments, demonstrating that visual reads performed in community settings are highly reliable.
The build-up of amyloid plaques in the brain is one of the first signs of Alzheimer's disease. Amyloid-specific PET tracers have been developed to identify this build-up at its earliest stages, revolutionizing how physicians identify and monitor the disease. Today, with the advent of novel disease-modifying therapies and widening reimbursement by third-party payers, amyloid PET is playing an increasing role in the diagnosis and management of patients with suspected Alzheimer's disease.
"For many years, amyloid PET was primarily used in research and clinical trials, and real-world experience was relatively limited," said Gil Rabinovici, MD, FAAN, FANA, professor of Neurology & Radiology at the University of California San Francisco. "As amyloid PET becomes more widely used, it's crucial to understand how reliably physicians can interpret the scans in everyday practice."
Rabinovici and colleagues utilized data from the Imaging Dementia—Evidence for Amyloid Scanning (IDEAS) study, which captured more than 18,000 clinical scan interpretations from local radiologists and nuclear medicine physicians in the United States. A subset of approximately 1,500 scans—consisting of 500 scans for each of the three U.S. Food and Drug Administration-approved tracers, 18F-florbetaben, 18F-florbetapir, and 18F-flutemetamol—were assigned to panels of three nuclear medicine or radiologist readers who are expert in interpreting amyloid PET scans (one panel per tracer) for visual interpretation. The agreement between the local readers and the expert panels was then compared.
Local readers interpreted 60.9 percent of scans as positive and 39.1 percent as negative. Across all radiotracers, agreement between majority expert and local reads was good, with 86.6 percent agreement for positive scans and 91.5 percent agreement for negative scans. Agreement did not vary significantly by PET tracer, and agreement between individual experts and local readers was also good.
"Our findings highlight the strong agreement between local and expert reads of scans," said Rabinovici. "They are very reassuring, demonstrating that patients and referring clinicians can have confidence in the interpretation of this high-stakes diagnostic test."
The authors of " Comparison of Expert Reads Versus Local Clinical Reads of Amyloid PET Scans in the IDEAS Study " include Konstantinos Chiotis, Edward and Pearl Fein Memory and Aging Center, Department of Neurology, Weill Institute for Neurosciences, University of California San Francisco, San Francisco, California, and Division of Clinical Geriatrics, Center for Alzheimer Research, Department of Neurobiology, Care Sciences and Society, Karolinska Institutet, Stockholm, Sweden; Charles C. Windon, Edward and Pearl Fein Memory and Aging Center, Department of Neurology, Weill Institute for Neurosciences, University of California San Francisco, San Francisco, California; Barry A. Siegel, Edward Mallinckrodt Institute of Radiology, Washington University School of Medicine, St. Louis, Missouri; Charles Apgar and Andrew Mach, Center for Research and Innovation, American College of Radiology, Reston, Virginia; Maria Carrillo, Medical & Scientific Relations Division, Alzheimer's Association, Chicago, Illinois; Ilana Gareen, Center for Biostatistics and Health Data Science, Brown University, Providence, Rhode Island, and Department of Epidemiology, Brown University School of Public Health, Providence, Rhode Island; Lucy Hanna and Justin Romanoff, Center for Biostatistics and Health Data Science, Brown University, Providence, Rhode Island; Constantine Gatsonis, Center for Biostatistics and Health Data Science, Brown University, Providence, Rhode Island, and Department of Biostatistics, Brown University School of Public Health, Providence, Rhode Island; and Gil D. Rabinovici, Edward and Pearl Fein Memory and Aging Center, Department of Neurology, Weill Institute for Neurosciences, University of California San Francisco, San Francisco, California, and Department of Radiology and Biomedical Imaging, University of California San Francisco, San Francisco, California.
Visit the JNM website for the latest research, follow our Twitter and Facebook
