Almost all magnetic resonance electrical impedance tomography (MREIT) reconstruction algorithms proposed to date assume isotropic conductivity in order to simplify the image reconstruction. However, it is well known that most of biological tissues have anisotropic conductivity values. In this study, four novel anisotropic conductivity reconstruction algorithms are proposed to reconstruct high resolution conductivity tensor images. Performances of these four algorithms and a previously proposed algorithm are evaluated in several aspects and compared.

Magnetic resonance electrical impedance tomography (MREIT) combines magnetic flux or current density measurements obtained by magnetic resonance imaging (MRI) and surface potential measurements to reconstruct images of true conductivity with high spatial resolution. Most of the biological tissues have anisotropic conductivity; therefore, anisotropy should be taken into account in conductivity image reconstruction. Almost all of the MREIT reconstruction algorithms proposed to date assume isotropic conductivity distribution. In this study, a novel MREIT image reconstruction algorithm is proposed to image anisotropic conductivity. Relative anisotropic conductivity values are reconstructed iteratively, using only current density measurements without any potential measurement. In order to obtain true conductivity values, only either one potential or conductivity measurement is sufficient to determine a scaling factor. The proposed technique is evaluated on simulated data for isotropic and anisotropic conductivity distributions, with and without measurement noise. Simulation results show that the images of both anisotropic and isotropic conductivity distributions can be reconstructed successfully.

Magnetic resonance conductivity tensor imaging (MRCTI) is an emerging modality which reconstructs images of anisotropic conductivity distribution within a volume conductor. Images are reconstructed based on magnetic flux density distribution induced by an externally applied probing current, together with a resultant surface potential value. The induced magnetic flux density distribution is measured using magnetic resonance current density imaging techniques. In this study, MRCTI data acquisition is experimentally implemented and anisotropic conductivity images of test phantoms are reconstructed using recently proposed MRCTI reconstruction algorithms.

Electrocardiography (ECG) signals and the information obtained through the analysis of these signals constitute the main source of diagnosis for many cardiovascular system diseases. Therefore, accurate analyses of ECG signals are very important for correct diagnosis. In this study, an ECG analysis toolbox together with a user-friendly graphical user interface, which contains the all ECG analysis steps between the recording unit and the statistical investigation, is developed. Furthermore, a new feature calculation methodology is proposed for ECG analysis, which carries distinct information than amplitudes and durations of ECG main waves and can be used in artificial intelligence studies. Developed toolbox is tested using both Massachusetts Institute of Technology-Beth Israel Hospital (MIT-BIH) Arrhythmia ECG Database and an experimentally collected dataset for performance evaluation. The results show that ECG analysis toolbox presented in this study increases the accuracy and reliability of the ECG main wave detection analysis, highly fasten the process duration compared to manual ones and the new feature set can be used as a new parameter for decision support systems about ECG based on artificial intelligence.

No abstract