Abstract:
|
This thesis consitutes the first development steps of PLAR visualization tool (theoretical design, properties and first tests). This tool aims to provide medical support in cardiologic field through the representation of the electric heart activity in a three-dimensional space. From electrocardiograms (ECGs), we have worked on a methodology to extract the data, filter it, process it and represent it in our PLAR space (which learns from most relevant directions in each patient). We tested our tool for different sets of ECGs, with patient having different cardiologic conditions, while tuning the parameters for the best visualization for each condition. |