Related Experiment Video
Updated: Oct 29, 2025
12:18
A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment
Published on: January 11, 2020
7.7K
A Real-Time Clinical Decision Support System, for Mild Cognitive Impairment Detection, Based on a Hybrid Neural
Carmen Paz Suárez-Araujo1, Patricio García Báez2, Ylermi Cabrera-León1
1Instituto Universitario de Ciencias y Tecnologías Cibernéticas, Universidad de Las Palmas de Gran Canaria, Las Palmas de Gran Canaria, Spain.
Computational and Mathematical Methods in Medicine
|July 14, 2021
Summary
Artificial neural networks (ANNs) can reliably estimate mild cognitive impairment (MCI) using common clinical data. This AI tool significantly outperforms physicians in diagnosing MCI, especially in primary care settings.
Area of Science:
- Artificial Intelligence in Medicine
- Neurology
- Geriatrics
Background:
- Diagnosing mild cognitive impairment (MCI) in primary care is challenging due to low suspicion and difficulties with cognitive test interpretation.
- Existing diagnostic methods for MCI often lack accuracy, particularly in non-specialist settings.
- Artificial neural networks (ANNs) offer potential for developing computer-aided diagnostic systems due to their pattern recognition and learning capabilities.
Purpose of the Study:
- To develop and evaluate a hybrid ANN-based system for reliable MCI estimation.
- To create a tool assisting clinical decision-making for MCI diagnosis using readily available primary care variables.
- To compare the diagnostic performance of the ANN system against that of clinical physicians.
Main Methods:
- An ANN model was designed using variables commonly available in primary care: Mini-Mental Status Examination (MMSE), Functional Assessment Questionnaire (FAQ), Geriatric Depression Scale (GDS), age, and years of education.
- Data from 128 MCI subjects and 203 controls were sourced from the Alzheimer's Disease Neuroimaging Initiative (ADNI).
- The ANN system identified optimal variable combinations, and its diagnostic accuracy (AUC, sensitivity, specificity, CUI) was compared with expert physician diagnoses.
Main Results:
- The optimal ANN model, utilizing MMSE, FAQ, and age, achieved an Area Under the Curve (AUC) of 95.2%, sensitivity of 90.0%, and specificity of 84.78%.
- Physician performance across a panel of specialists (family physicians, neurologist, geriatrician) yielded a sensitivity of 46.66% and specificity of 91.3%.
- The ANN model demonstrated superior clinical utility index (CUI) compared to physician performance.
Conclusions:
- The proposed ANN system achieves excellent diagnostic accuracy for MCI, even when relying solely on common clinical tests.
- The system's high performance and reliance on standard data make it highly suitable for implementation in primary care settings.
- This AI-driven tool can effectively aid primary care physicians in identifying and managing suspected cases of cognitive impairment.
