Machine learning approaches with cognitive data for early detection of dementia
: a scoping review

  • Raquel Gonçalves Colaço (Student)

Student thesis: Master's Thesis


Objective: The objective of this scoping review is to map the literature on the types of data commonly used in AI models for the early detection of Alzheimer’s Disease (AD). Particularly, we will look for information about the extent to which cognitive data is considered in this type of research. Introduction: Alzheimer’s Disease is the most common form of dementia and the seventh leading cause of death worldwide. Dementia has massive individual, societal and economic impact and because there is no approved cure or modifying therapies available yet, the investigation should focus on early detection of AD and dementia prevention. Inclusion Criteria: Studies from 2017 until March 2023 were included. The main inclusion criteria included studies that focused their investigation specifically on Alzheimer’s disease and aimed to develop/train AI models to assist on early-detection of the disease and on the prevention of dementia. Methods: This scoping review was developed based on the JBI methodology for scoping reviews (Peters et al., 2020) and conducted on Medline Complete via EBSCO, PubMed and Web of Science on March 2023. Articles included in this review were screened by two independent reviewers and after being analysed, the relevant information was retrieved. Results: All articles have taken into account either cognitively normal (CN) individuals, mild cognitive impairment (MCI) individuals or Alzheimer’s disease (AD) individuals, and trained models based on machine learning, the most common type of data used to train models was brain MRI, cognitive data is mainly used to characterize the cognitive status of the participants (to form the experimental and control groups) and the majority of these investigations are published in engineering journals. Conclusions: Neuroimage is the type of data where investigation in this area is more focused on, and there is a need to include novel types of data that are also relevant to the early detection of Alzheimer’s Disease, namely cognitive data. It is also important to x make this type of information more easily available for psychologists as well as educate them on this subject since such technology can bring great benefits to their practice.
Date of Award11 Dec 2023
Original languageEnglish
Awarding Institution
  • Universidade Católica Portuguesa
SupervisorMaria Vânia Silva Nunes (Supervisor) & Spyros Angelopoulos (Co-Supervisor)


  • Machine learning
  • Early detection of AD
  • Prevention of dementia
  • Cognitive data
  • Neuroimagem


  • Mestrado em Neuropsicologia

Cite this