Screening for Alzheimer's disease in the community using an AI-driven screening platform: design of the PREDICTOM study.
Publisher DOI
PubMed ID
41894949
Abstract
Background
Recent developments in physiological, imaging and digital biomarkers combined with the approval of new disease-modifying drugs against Alzheimer's disease (AD) and diagnostic blood tests provide an opportunity to shift the first diagnostic steps to the home-setting. While these novel biomarkers enable scalable screening and earlier detection and treatment of AD, they require an evaluation of their accuracy, feasibility, and safety in primary care and the community setting.
Objectives
The aim of PREDICTOM is to develop and test the accuracy of an artificial intelligence (AI) driven screening platform for the risk assessment and early detection of AD to extend the clinical pathway to home-based screening using established and novel biomarkers.
Design/setting
PREDICTOM is a European (Norway, UK, Belgium, France, Switzerland, Germany, Spain) observational, prospective cohort study using a cloud-based platform that stores a digitalised journey for each participant and provides a collection of artificial-intelligence (AI) algorithms and tools for risk assessment and early diagnosis and prognosis.
Participants
Cohort 1 consists of 4000 adults aged 50 years or older at risk of developing AD. Cohort 2 consists of 615 participants selected from Cohort 1 based on estimates indicating high (N = 415) or low (N = 200) risk of AD. Data from existing cohorts will guide the analytic strategy of the study.
Measurements
Cohort 1 will undergo home-based assessments (Level 1), Cohort 2 will undergo in-clinic assessments (Levels 2 and 3). Level 1 includes at-home screening, collecting digital and physiological data (questionnaires, cognition, hearing, eye-tracking) and biofluids (capillary blood via finger-stick and saliva) for biomarker analysis. Level 2 comprises a more complex biomarker collection, most of which can be completed in primary care, including EEG, MRI, venous blood, microbiome from stool, cognition, hearing, and eye-tracking. Level 3 includes a diagnostic evaluation to confirm or rule out AD pathology using established biomarkers (cerebrospinal fluid, or amyloid PET).
Conclusions
PREDICTOM will develop AI-driven algorithms for the early detection of AD using biomarkers that can be collected at home or in the community care setting, and evaluate their integration into a well-defined and comprehensive clinical pathway.
Recent developments in physiological, imaging and digital biomarkers combined with the approval of new disease-modifying drugs against Alzheimer's disease (AD) and diagnostic blood tests provide an opportunity to shift the first diagnostic steps to the home-setting. While these novel biomarkers enable scalable screening and earlier detection and treatment of AD, they require an evaluation of their accuracy, feasibility, and safety in primary care and the community setting.
Objectives
The aim of PREDICTOM is to develop and test the accuracy of an artificial intelligence (AI) driven screening platform for the risk assessment and early detection of AD to extend the clinical pathway to home-based screening using established and novel biomarkers.
Design/setting
PREDICTOM is a European (Norway, UK, Belgium, France, Switzerland, Germany, Spain) observational, prospective cohort study using a cloud-based platform that stores a digitalised journey for each participant and provides a collection of artificial-intelligence (AI) algorithms and tools for risk assessment and early diagnosis and prognosis.
Participants
Cohort 1 consists of 4000 adults aged 50 years or older at risk of developing AD. Cohort 2 consists of 615 participants selected from Cohort 1 based on estimates indicating high (N = 415) or low (N = 200) risk of AD. Data from existing cohorts will guide the analytic strategy of the study.
Measurements
Cohort 1 will undergo home-based assessments (Level 1), Cohort 2 will undergo in-clinic assessments (Levels 2 and 3). Level 1 includes at-home screening, collecting digital and physiological data (questionnaires, cognition, hearing, eye-tracking) and biofluids (capillary blood via finger-stick and saliva) for biomarker analysis. Level 2 comprises a more complex biomarker collection, most of which can be completed in primary care, including EEG, MRI, venous blood, microbiome from stool, cognition, hearing, and eye-tracking. Level 3 includes a diagnostic evaluation to confirm or rule out AD pathology using established biomarkers (cerebrospinal fluid, or amyloid PET).
Conclusions
PREDICTOM will develop AI-driven algorithms for the early detection of AD using biomarkers that can be collected at home or in the community care setting, and evaluate their integration into a well-defined and comprehensive clinical pathway.
Date Issued
2026
Publication Type
Article
Subject(s)
Subjects
Alzheimer’s disease
•
Artificial intelligence
•
Biomarker
•
Early detection
Language(s)
en
Author(s)
Khan, Zunera | |
Radermacher, Jonas | |
Georgiadis, Kostas | |
Lazarou, Ioulietta | |
Grammatikopoulou, Margarita | |
Pickering, Ellie | |
Mitterreiter, Johanna | |
Aakre, Jon Arild | |
Ashton, Nicholas J | |
Baquero, Miguel | |
Beser-Robles, Maria | |
Braboszcz, Claire | |
Brandt, Sigurd | |
Brown, James | |
Cacciamani, Federica | |
Campill, Sarah | |
Collins, Christopher | |
Deshpande, Pushkar | |
Diaz, Ana | |
Durrleman, Stanley | |
Engelborghs, Sebastiaan | |
Ferré-González, Laura | |
Frisoni, Giovani B | |
Gjestsen, Martha Therese | |
Gove, Dianne | |
Honigberg, Lee | |
Huang, Bin | |
Hudak, Anett | |
Kaushik, Sandeep | |
Letoha, Tamas | |
Marquardt, Gaby | |
Mendes, Augusto J | |
Müllenborn, Matthias | |
Paletta, Lucas | |
de Barros, Nuno Pedrosa | |
Pszeida, Martin | |
Vik-Mo, Audun Osland | |
Rostamipour, Hossein | |
Perneczky, Robert | |
Rauchmann, Boris-Stephan | |
Russegger, Silvia | |
Schirmer, Timo | |
Shadmaan, Amied | |
Solana, Ana Beatriz | |
Soria-Frisch, Aureli | |
Tegethoff, Paulina | |
Ribbens, Annemie | |
De Witte, Sara | |
van der Giezen, Mark | |
Nikolopoulos, Spiros | |
Corbett, Anne | |
Fröhlich, Holger | |
Aarsland, Dag |
Additional Credits
Journal
The Journal of Prevention of Alzheimer’s Disease
Publisher
Elsevier
ISSN
2426-0266
2274-5807
Access(Rights)
open.access