Implementation of an artificial neural network on the test Barcelona workstation as a predictive model for the classification of normal, mild cognitive impairment and Alzheimer's disease subjects using the Neuronorma battery

Objective: to implement an online Artificial Neural Network (ANN) that provides the probability of a subject having mild cognitive impairment(MCI) or Alzheimer’s disease (AD). Method: Different ANN’s were trained with a sample of 350 controls, 75 MCI and 93 AD subjects. The ANN structure chosen was:...

Descripción completa

Detalles Bibliográficos
Autores: Rivera Avila, Neus, Cabrera-Bean, Margarita|||0000-0001-9058-3825, Sánchez Benavides, Gonzalo, Gallego González, Carles, Lupiáñez Pretel, José, Peña Casanovas, Jordi
Tipo de recurso: artículo
Fecha de publicación:2018
País:España
Institución:Universitat Politècnica de Catalunya (UPC)
Repositorio:UPCommons. Portal del coneixement obert de la UPC
Idioma:inglés
OAI Identifier:oai:upcommons.upc.edu:2117/125203
Acceso en línea:https://hdl.handle.net/2117/125203
https://dx.doi.org/10.18502/kls.v4i8.3334
Access Level:acceso abierto
Palabra clave:Machine learning
Neural networks (Computer science)
Neural networks
Cognitive impairment
Alzheimer’s disease
NEURONORMA battery
Aprenentatge automàtic
Xarxes neuronals (Informàtica)
Àrees temàtiques de la UPC::Informàtica::Intel·ligència artificial::Aprenentatge automàtic
Descripción
Sumario:Objective: to implement an online Artificial Neural Network (ANN) that provides the probability of a subject having mild cognitive impairment(MCI) or Alzheimer’s disease (AD). Method: Different ANN’s were trained with a sample of 350 controls, 75 MCI and 93 AD subjects. The ANN structure chosen was: (1) an input layer of 33 cognitive variables from the NEURONOR-MA battery plus two sociodemographic variables, age and education, which was reduced to a 15-fea-ture input vector using the Multiple Discriminant Analysis method; (2) one hidden layer with 8 neurons; and (3) three outputs corresponding to the 3cognitive states. This ANN was determinate in a previous study. The ANN was implemented in TestBarcelona Workstation.189 Results: The best designed ANN diagnoses with a probability of 94.87% subjects well classified when comparing controls, MCI and AD using the NEURONORMA battery. Conclusions: ANNs are a powerful tool for classifying subjects with cognitive impairment using the NEURONORMA battery. When a single profile is entered, it delivers the probabilities of be-longing to each one of the three cognitive states. This constitutes a good complement to the interpretation of neuropsychological profiles for clinical decision making.