"Methoden zur Bestimmung von dreidimensionalen intrakardialen und extrakorporalen Elektrodenpositionen"
Verlagshaus Monsenstein und Vannerdat, ISBN 3-935363-43-5, Aug. 2001
(english abstract - pdf)

Introduction   Analyzing the electrical activity is of importance in the field of cardiac diagnosis and surgical treatment planning. A significant aspect deals with the mapping of electrical endocardial potentials. This kind of measurement is accomplished with catheters which have a certain amount of electrodes. In the past few years the amount of such systems increased in the clinical practice. In literature different systems are described. These systems have local coordinate systems, like the catheter itself.

Surface measurement and mapping of extracorporal potentials is known for some time. An electrocardiogram (ECG) is recorded with few electrodes (e.g. three). The usage of e.g. 32 electrodes leads to so called body surface potential maps (BSPM). The combination of extracorporal and intracardial potential maps establish new possibilities to obtain more information about the electrical activity of the heart.

For a computer simulation of the physiology of the heart and for a validation of generated heart models the electrical measurement, a conductivity model of the investigated body and the global three-dimensional measurement positions are needed. A conductivity model can be built from computer tomography (CT) or magnetic resonance imaging (MRI). In this work the three-dimensional extracorporal electrode positions are determined with a camera system. The intracardial electrode positions on the catheters are acquired from the tomographic images. The combination of such generated computer models leads to hybrid data sets including intracardial and extracorporal electrode positions and a volume data set. These hybrid data sets are used for numerical field calculation, data analysis and data validation. Examples are the inverse and the forward problem of electrocardiography.

Conceptual formulation   The compilation of a complete set of electrical and morphological data requires the three-dimensional localization of electrodes in the heart as well as on the thorax and the segmentation and classification of three-dimensional volume data sets.

For the calculation of the extracorporal positions a camera system is developed and validated. Therefore, reference objects are designed which allow a calibration and validation of such a camera system.

In the scope of this work, various types of catheters must be extracted from different image modalities and models generated. Methods are developed for the different image modalities to localize the intracardial electrode positions on these catheters. Again, reference objects are built to calibrate and validate the X-ray system.

A part of this work is embedded into the german Sonderforschungsbereich 414: 'Informationstechnik in der Medizin - Rechner- und sensorgestützte Chirurgie' (Special research area 414: 'Information technology in medicine - computer and sensor aided surgery'). It is a joint project of the Universities Karlsruhe and Heidelberg and the German Cancer Research Institute. One part of this project is 'H4: Elektromechanische Modellierung des Vorhofes: Validierung, Planung und Simulation atrialer rythmuschirurgischer Eingriffe' ('H4: Electro-mechanical modeling of the atrium: Validation, planning and simulation of atrial rhythm surgical intervention'). Its aim is among other things to simulate the electro-mechanical behaviour of the atrium. Within this project, high-spatial resolution macroscopic anatomic heart models are created from animal experiments. Morphological computer tomographic images and electrical data sets are acquired and visualized.

The calculated intracardial and extracorporal electrode positions are transformed in one global coordinate system and into a three-dimensional volume data set afterwards. These three-dimensional volume data sets are segmented and classified from MRI and CT images.

Structure and Methods   The first part of this work introduces the basics from mathematics, digital imaging and the camera systems. The mathematical basics include different transformations, the eigenvalue theory and linear equations, the Hessian Matrix, interpolation methods and the extremum determination of multi-dimensional functions. The digital imaging section gives an overview about image filtering and image segmentation. Image filtering allows certain information to be stressed and other to be suppressed. In general, there are linear, non-linear and hybrid filters. Image segmentation intersects the image into homogenous regions by using point, edge or regional methods. The camera system section shows the different optical mapping features of a video and a X-ray system and also the aberration correction.

The basics of the anatomy, physiology and pathology are described in the next chapter of this work. The current state of research in the field of "Mapping of potential distribution in the heart" is outlined and an overview about the electrode catheters is given. In general, one can distinguish between three types of catheters: String, basket and balloon catheters. String catheters have n electrodes in a row on one string. Basket catheters have a set of strings (usually eight) which have a certain amount of electrodes (n=8). This leads to a total amount of 64 electrodes for a basket catheter. The balloon catheter has also 64 electrodes. Here, the electrodes are on a wire grid. A balloon inside this wire grid can be filled with a fluid (normally some contrast agent) to expand the balloon. The string catheters serve for measurement and ablation, whereas the other two only serve for measurement.

Another part of this work introduces the different calibration techniques of the imaging systems and the methods for the determination of the three-dimensional electrode positions. Various reference objects are used for the calibration of the camera and the X-Ray system. A calibration matrix for each camera and each X-Ray tube describes the image mapping of the camera respectively the X-Ray tube. Afterwards, a three-dimensional electrode position is calculated from two different images by solving an overdetermined equation system. The image position of the electrodes and the three-dimensional calibration values build this equation system. The image positions of the electrodes in the camera and the X-Ray system are extracted with image filtering and segmentation methods. The rigid transformation of the electrodes in a global coordinate system and also into the volume data set is accomplish by minimizing a quality function. The distances between the electrodes respectively the distances between the electrodes an the thorax surface build this quality function. A computer model of the balloon catheter is generated for the registration of the electrode positions in the three-dimensional computer tomography images.

The results of this work are presented in a next part. The camera and X-Ray system are calibrated and validated with reference objects. The localization of the extracorporal electrodes on a human thorax is presented. The three-dimensional positions of different types of catheters are calculated from the X-Ray and CT system.

One section describes the transformation of the electrode positions from the camera and X-Ray system into a unitary coordinate system and into a three-dimensional volume data set for a test-setup.

In the context of the animal experiments the modeled balloon catheter is segmented from the morphological CT images and positioned into the three-dimensional segmented animal heart. Afterwards, the measured electrical data is projected onto the catheter.

Morphological MRI images are taken from a patient in a clinical study. The patient has a string catheter ablation in the heart with the aid of the measurement of a basket catheter. X-Ray images are taken and electrical data is collected. The camera system takes images from the patient's thorax where the extracorporal electrodes are fixed. The three-dimensional intracardial and extracorporal electrode positions are calculated and positioned into the three-dimensional volume data set of the patient.

Summary   To summarize this work it can be stated, that this work presents new possibilities to generate a complete set of morphological and electrical data in combination with the three-dimensional measurement position. With this ability it is possible in future times to validate computer simulations of the electrical excitation propagation and also to accommodate the measurement data of an individual patient for development of an optimal therapy strategy.