Neuroimaging‐based artificial intelligence for predicting dementia progression in a diverse real‐world memory clinic cohort
Models trained on large, research cohorts are effective at predicting progression to dementia for real-world NHS patients, despite differences in the sample characteristics. p<sub>AD</sub> was significant at estimating time-to-diagnosis, with a 0.1 increase in p<sub>AD</sub> equating to a 13.6%-16.7% increased risk of receiving a dementia diagnosis within 8 years. These findings demonstrate the clinical utility of generalisable artificial intelligence models in a real-world context.
Authors
- Amelia Jewell (ORCID: https://orcid.org/0000-0002-0887-2159)
- Francesca Biondo (ORCID: https://orcid.org/0000-0001-9952-0249)
- Sophie Martin (ORCID: https://orcid.org/0000-0003-3819-1634)
- Frederik Barkhof
- James H. Cole
Institutions
- King's College London (GB)
- South London and Maudsley NHS Foundation Trust (GB)
- The London College (GB)
- National Hospital for Neurology and Neurosurgery (GB)
- University College London (GB)
- Amsterdam UMC Location Vrije Universiteit Amsterdam (NL)
Publication Details
- Journal
- Alzheimer s & Dementia
- Published
- 2025-12-01
- DOI
- https://doi.org/10.1002/alz70856_102951
- Primary Topic
- Dementia and Cognitive Impairment Research
- Type
- article
- Field-Weighted Citation Impact
- 0.00