Computational modelling of deficits in attentional networks in mild traumatic brain injury: an application in neuropsychology

This work builds upon a model of performance for the Attentional Network Test (ANT) implemented in ACT-R 6.0 (Hussain & Wood, in press; 2008) to simulate neglect conditions related to mild Traumatic Brain Injury (mTBI) and their effect on the attentional networks of alerting, orienting and executive control (cf. Posner & Fan, 2007). The model is evaluated against data sets for a human study of recovery in mTBI patients (Halterman, Langan, Drew, Rodriguez, Osternig, Chou & Donkelaar, 2006) and the model’s fit to data is assessed statistically. A description of the mTBI model is provided, outlining how the simulation of impairment is achieved using ACT-R’s1 symbolic and subsymbolic components. The model data further indicates that modulation effects previously observed between attentional networks (Hussain & Wood, in press; 2008) are preserved in the mTBI model, and do not reflect the impairment observed in orienting and executive control efficiency. The exercise of model fitting and the model code itself further establishes that alerting remains unimpaired in mTBI and the orienting network impairment is be mainly due to affected disengaging ability from invalidly cued location to target location.