Determining the multi-current sources of magnetoencephalography by using Fuzzy Topographic Topological Mapping

Abstract

The human brain is an extremely complex system performing demanding information processing tasks rapidly. It consists of billions of neurons, each connected to others through thousands of synapses or interconnections. This huge network has many electric and chemical processes that can be measured in various ways. Magnetoencephalography (MEG) is a technique of measuring and recording the minute and very weak magnetic fields generated by the currents in the neurons. There are two types of problems in MEG, the forward problem and the backward or the inverse problem. The forward problem deals with finding the magnetic fields when the current source distribution is given or known. On the other hand, the inverse problem is to find the neural current source distribution given a series of magnetic fields measurements. This study has proposed the model FTTM2 (Fuzzy Topographic Topological Mapping Version 2) which is an extension to the novel mathematical modeling FTTM1 (Fuzzy Topographic Topological Mapping Version 1). The model FTTM2 comprises four components namely the Image Contour Plane (IC), Base Image Plane (BI), Fuzzy Image Field (FI) and Topographic Image Field (TI). In the process of applying FTTM2, emphasis is made on its first component, the IC where two different algorithms are being applied to the data. The first is the fuzzy c-means (FCM) algorithm which is used to determine the region where the current sources lie and also to approximate the number of current sources. The second is the seed-based region growing (SBRG) algorithm which is used to confirm the number of current sources available in the system by automation. Two theorems and three corollaries are derived and proven as theoretical basis of the proposed system. Finally, FTTM2 is tested on the generated and experimental data and subsequently verified using forward and backward calculations