Background

The PREDICTOM project (Prediction of Neurodegenerative Disease using an AI-driven Screening Platform, www.predictom.eu) is an EU-funded initiative aimed at developing novel biomarkers to identify individuals at high risk of Alzheimer’s disease (AD)¹. This work describes the rationale and design of an advanced MRI protocol implemented at four of seven PREDICTOM clinical centers, complementing the harmonized state-of-the-art protocol².

Methods

The study will acquire MRI data from 615 participants (target ratio: 80% higher risk / 20% lower risk), all healthy and aged ≥50 years. Seven European centers equipped with 3T scanners from two vendors and six models will collect harmonized state-of-the-art MRI data².
In addition, four centers using GE HealthCare 3T systems will implement an advanced MRI protocol evolving from conventional sequences (T1, T2, T2-FLAIR, SWI/T2*, DWI) and advanced scans (resting-state fMRI, ASL, MRS) to a next-generation workflow featuring: Highly accelerated T1/T2/T2-FLAIR with deep-learning reconstruction³,  improved SWI/T2 via 3D EPI⁴*, multi-echo resting-state fMRI⁵, enhanced ASL with seven PLDs⁶, multi-contrast quantitative mapping⁷ and a novel water-exchange contrast⁸.

Results

Figure 1 illustrates representative images and acquisition times from a healthy volunteer scanned at a PREDICTOM site. Existing and novel imaging biomarkers, such as volumetry, cortical thickness, white-matter hyperintensity, microbleeds or connectivity measures will be derived from these faster, higher-SNR, and more accurate sequences and compared against standard protocol outputs. Additionally, the value of quantitative maps and novel biomarkers in early detection of AD will be also evaluated in the PREDICTOM multi-center cohort.

Conclusion

The advanced PREDICTOM MRI protocol enables validation of next-generation imaging techniques to extract both established and novel biomarkers, supporting early identification of individuals at high risk for AD

(1) Brem, AAIC2024, (2) Fernandez, OHBM2025, {3} Ahn, ISMRM2023,(4) Huang, MRM2024, (5) Fernandez, Neuroimage2017 (6) van der Thiel, AJNR2018, (7) Ljungberg E, PNMRS2021, (8) Taso, MRM2025

Keywords: MRI, early detection, Alzheimer’s Disease 

Learning Objective: Towards next generation MRI protocols for early AD prediction