Noninvasive source imaging from ECG mapping data

The research in P14207-MED deals with the modelling of the human thorax and with a new mathe-matical procedure to the reconstruction of the electric activity in the human heart. The developed methods were successfully applied to four patients and clinically validated. So far, the diagnostics of cardiac arrhythmias takes place with catheter-mapping or with conventional catheter techniques. These interventions represent a serious risk for the patient, do not allow the detection of "single events" and, moreover, can not be used for screening. For this reason world-wide efforts have already been under-taken to develop a non-invasive method with which above disadvantages can be recovered. In the project P14207-MED research and development has been advanced enabling such a non-invasive di-agnostics procedure. First of all, anatomical data (e.g., be means of magnetic resonance (MR) imag-ing) and an extended electrocardiogram (ECG mapping, e.g., with 65 electrodes) are acquired from the patient. From the MR data a computer model of the thorax (volume conductor model) is generated, which describes the physical relationship between the electrical sources in the cardiac muscle and the ECG mapping data. Based in this computer model, afterwards from the ECG mapping data, by the solution of a so- called ill-conditioned inverse problem, the electrical activation sequence (i.e., the spa-tio-temporal propagation of the electrical excitation) in the cardiac muscle is calculated and graphi-cally displayed. With this new approach the same information can be non-invasively supplied to the cardiologist as it is possible with invasive methods. The solution of this inverse problem under clinical conditions represents a very difficult step and can be regarded as substantial innovation. In P14207-MED two technically-scientific aspects of detail were particularly treated: 1.) the mathematical model of the volume conductor: A new triangular formulation was developed allowing individual thorax ge-ometry modelling to be performed very exactly and 2.) a new reconstruction algorithm for the calcula- tion of the electrical activation sequence. These two technical innovations were tested on the basis of four patient measurements and clinically evaluated. The obtained localization results (in the compari-son with the clinical reference method) were for sinus rhythm and for focal pacing in the right ventri-cle within the range of 6 to 14 millimeters. For the quantitative evaluation the first endocardial break-through of the electrical excitation was considered. These and other research is further investigated intensively in the START Y144-INF program. Global goal of these research is to extend the anatomi-cal imaging procedures such as MR, ultrasound, computer tomography, and biplane-fluoroscopy by the imaging of electrical function in the cardiac muscle. Additionally, the area of inverse electrocar-diography is to be developed for a potential application and clinically established. On a long-term ba-sis seen the non-invasive imaging of electrical function in the human heart will have extensive effects on the diagnostics and therapy of cardiac arrhythmias.