Source clustering for forward problem of FCG

Abstract

This thesis presents the minimum numbers of location clusters optimized for various factors on which the results are dependent; i.e., ECG leads, heart and torso (volume conductor) models, and desired accuracies of the forward solution. Spatial partitions and cluster sizes of the optimized clusters are also presented and discussed. Each cluster indicates a region in which source-to ECG measurement transformation can be considered approximately identical. The results suggest that for higher accuracy in the forward solutions, a larger number of clusters and thus a higher spatial resolution is required. The analysis, results and discussion presented in this thesis are not only valid in the field of ECG simulations, but also in various other fields of study such as source-to-measurement relationship analysis and future improvement of bioelectric measurement methods, giving this work a broad perspective.

AB3:Tavoitteena oli tutkia sydämen sähköisten pulssien lähteiden (noodien) sähkötoiminnan ja kehon pinnalta mitatun sähköisen toiminnan välistä matemaattista suhdetta. Työssä on analysoitu tätä sekä kehitetty malli, jolla voidaan yksinkertaisemmin laskea sydämen toiminnan aiheuttama kehon pinnan potentiaalijakauma klusteroimalla samankaltaisia lähteitä yhteen.