A predictive personalised model for the left atrium

Abstract

We propose an integrated modelling and clinical protocol for characterising local tissue properties in the left atrium and validate the resulting personalised models in four patient cases. We generate a personalised model from a set of measured local activation times (LATs) obtained by pacing the left atrium in the proximity of the coronary sinus with a programmed pacing protocol. We validate the model by evaluating the correlation between a set of measured LATS, obtained by pacing on the high right atrium and a set numerically computed LATs. We then estimate if the tissue is capable of sustaining an atrial fibrillation or a tachycardia by triggering a spiral wave on the computational model and then analysing the activation frequencies and the time elapsed until the termination of the aberrant activation.