Generation of Synthetic [18F]FDG PET From Early-Phase Amyloid PET in Alzheimer's Disease

This study aims to test a new artificial intelligence (AI) method to create brain scan images without needing an extra scan. Currently, patients with memory problems often undergo two types of PET scans (Amyloid PET and FDG PET) to assess Alzheimer's disease. This study will use existing scan data from patients who already had both scans as part of their routine care.

The AI model will try to generate the FDG PET image using only the Amyloid PET scan and an MRI. If successful, this method could reduce radiation exposure, costs, and time for future patients by eliminating the need for a separate FDG injection and scan.

No new scans, injections, or procedures will be performed for this study. All data will be fully anonymized (personal information removed) before analysis. The study involves approximately 35 adult patients (age 50+) whose data were collected between January 2025 and December 2025 at IRCCS Ospedale San Raffaele in Milan, Italy.

Study Overview

• Status: Not yet recruiting
• Conditions: Alzheimer Disease (AD)

Detailed Description

This is a retrospective observational study conducted at IRCCS Ospedale San Raffaele, Milan, Italy. The study evaluates the accuracy of synthetic [18F]FDG PET images generated using a SwinUNETR deep learning model compared to native [18F]FDG PET images.

Study Population:

Adults (≥ 50 years) who underwent amyloid PET imaging (using Florbetaben or Flutemetamol), structural MRI, and [18F]FDG PET due to cognitive symptoms between January 2025 and December 2025. Approximately 35 patients meeting inclusion criteria will be included.

Methodology:

All imaging and clinical data were collected as part of routine diagnostic care; thus, no additional procedures, interventions, or interactions with patients are required for this study. All data are fully deidentified before analysis, consistent with GDPR and institutional data protection policy. The SwinUNETR model processes volumetric images to generate synthetic FDG PET images from early-phase amyloid PET and MRI inputs.

Objectives and Endpoints:

Primary Objective: To quantitatively and qualitatively assess the accuracy of synthetic FDG PET images compared with native FDG PET images.

Primary Endpoint: Pearson correlation coefficient and mean absolute error (MAE) of SUVR values obtained from native FDG PET and synthetic FDG PET in Alzheimer relevant areas of interest (precuneus, posterior cingulate, lateral temporal cortex, and frontal cortex).

Secondary Objective: To assess visual interpretability and clinical intuitiveness of synthetic FDG PET images by expert nuclear medicine physicians.

Secondary Endpoint: Inter-rater agreement (Cohen's kappa) among 2 blinded nuclear medicine physicians rating synthetic FDG scans as "clinically acceptable" or not.

Ethical Considerations:

Due to the retrospective and non-interventional nature of this study, no additional informed consent is required. A waiver of informed consent will be requested from the Ethics Committee. The image data will be fully anonymized in accordance with institutional policies.

• Study Type: Observational
• Enrollment (Estimated): 35

Contacts and Locations

• Study Contact: Name: Clinical Trial Center Nuclear Medicine Unit; Phone Number: +39-02-2643-2716; Email: mednuc@hsr.it
• Study Contact Backup: Name: CTC First Contact ctc.firstcontact@hsr.it

Study Locations

• Italy
  • Lombardy
    • Milan, Lombardy, Italy, 20132
      • Irccs Ospedale San Raffaele

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Adult; Older Adult
• Accepts Healthy Volunteers: No

• Sampling Method: Non-Probability Sample

Study Population

Adults (≥50 years) with cognitive symptoms who underwent amyloid PET imaging (Florbetaben or Flutemetamol), structural MRI, and [18F]FDG PET at the Nuclear Medicine Department of IRCCS Ospedale San Raffaele, Milan, Italy, as part of routine diagnostic care between January 1, 2025 and December 31, 2025.

Description

Inclusion Criteria:

• Age ≥ 50 years at the time of imaging.
• Clinically indicated amyloid PET scan performed with Florbetaben or Flutemetamol between January 1, 2025 and December 31, 2025.
• Availability of paired structural MRI (3D T1-weighted) and real [18F]FDG PET scan acquired within ±6 months of the amyloid PET.
• All three imaging modalities (Amyloid PET, FDG PET, MRI) are of sufficient technical quality for co-registration and quantitative analysis.

Exclusion Criteria:

• Presence of other major neurological disorders that may confound FDG metabolism (e.g., Parkinson's disease, frontotemporal dementia, brain tumor, or recent stroke).
• Severe motion artifacts or technical failures in any of the three imaging modalities that prevent reliable co-registration or SUVR calculation.
• Incomplete or irreversibly corrupted DICOM data preventing anonymization or conversion to analysis-ready format.

Study Plan

How is the study designed?

Number of groups / cohorts

1

Cohorts and Interventions

Group / Cohort
Patients with cognitive impairment undergoing amyloid PET; Adults (≥50 years) with cognitive symptoms who underwent amyloid PET (Florbetaben or Flutemetamol), structural MRI, and [18F]FDG PET at IRCCS Ospedale San Raffaele between January 2025 and December 2025 as part of routine diagnostic care. All data are fully anonymized prior to analysis.

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Quantitative Accuracy of Synthetic FDG PET Images (SUVR Correlation and MAE)	Pearson correlation coefficient and mean absolute error (MAE) of SUVR values obtained from native FDG PET and synthetic FDG PET in Alzheimer relevant areas of interest (precuneus, posterior cingulate, lateral temporal cortex, and frontal cortex).	Retrospective analysis of imaging data acquired between January 1, 2025 and December 31, 2025

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Regional SUVR Bias Between Synthetic and Native FDG Across Machine Types	Regional SUVR bias (mean difference ± standard deviation) between synthetic and native FDG across machine types and reconstructions used at Ospedale San Raffaele.	Retrospective analysis of imaging data acquired between January 1, 2025 and December 31, 2025

Other Outcome Measures

Outcome Measure	Measure Description	Time Frame
Effect of Amyloid Status on Synthetic FDG Generation Accuracy	Stratified analysis of primary endpoint (SUVR correlation and MAE) by state of amyloid PET positivity (positive vs negative) based on established Centiloid or visual read criteria.	Retrospective analysis of imaging data acquired between January 1, 2025 and December 31, 2025

Collaborators and Investigators

• Sponsor: IRCCS San Raffaele

Study record dates

Study Major Dates

• Study Start (Estimated): March 1, 2026
• Primary Completion (Estimated): December 31, 2026
• Study Completion (Estimated): December 31, 2026

Study Registration Dates

• First Submitted: February 18, 2026
• First Submitted That Met QC Criteria: February 18, 2026
• First Posted (Actual): February 24, 2026

Study Record Updates

• Last Update Posted (Actual): February 24, 2026
• Last Update Submitted That Met QC Criteria: February 18, 2026
• Last Verified: February 1, 2026

More Information

Terms related to this study

• Keywords: Positron Emission Tomography; Alzheimer Disease; Artificial Intelligence; Neuroimaging; Deep Learning; FDG PET; Diagnostic Imaging; Image Processing; Retrospective Study; Flutemetamol; Florbetaben; Amyloid PET
• Additional Relevant MeSH Terms: Brain Diseases; Central Nervous System Diseases; Nervous System Diseases; Mental Disorders; Neurocognitive Disorders; Dementia; Tauopathies; Neurodegenerative Diseases; Alzheimer Disease
• Other Study ID Numbers: ALZ-AI

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: UNDECIDED
• IPD Plan Description: It is not yet decided whether individual participant data will be shared. The research team will consider data sharing requests on a case-by-case basis, subject to institutional approval and data use agreements. Anonymized imaging data (NIfTI format) and analysis scripts may be made available upon reasonable request to the Principal Investigator.

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No