Messinger-Rapport B J, Rudy Y
Department of Biomedical Engineering, Case Western Reserve University, Cleveland, Ohio 44106.
Circ Res. 1990 Apr;66(4):1023-39. doi: 10.1161/01.res.66.4.1023.
The inverse problem in electrocardiography implies the reconstruction of electrical events within the heart from information measured noninvasively on the body surface. Deduction of these electrical events is possible from measured epicardial potentials, and, thus, a noninvasive method of recovering epicardial potentials from body surface data is useful in experimental and clinical studies. In the present study, an inverse method that uses Tikhonov regularization was shown to reconstruct, with good accuracy, important events in cardiac excitation. The inverse procedure was employed on data obtained from a human-torso tank in which a beating canine heart was placed in the correct anatomical position. Comparison with the actual, measured epicardial potentials indicates that positions and shapes of potential features (maxima, minima, zero potential line, saddles, etc.) are recovered with good accuracy throughout the QRS. An error in position of up to 1 cm is typical, while amplitudes are slightly diminished. In addition, application was extended from the above setting, in which the geometry was precisely known and potentials at a large number of leads were measured accurately, to a situation that is more representative of clinical and experimental settings. Effects of inaccuracy in location of the position of the heart were examined. A stylized torso that approximates the actual geometry was designed, and its performance in the inverse computations was evaluated. A systematic method of reduction of the number of leads on the body surface was proposed, and the resulting lead configurations were evaluated in terms of the accuracy of inverse solutions. The results indicate that the inverse problem can be stabilized with respect to different types of uncertainties in measured data and offer promise in the use of the inverse procedure in clinical and experimental situations.
心电图中的逆问题意味着从体表无创测量的信息中重建心脏内部的电活动。从测量的心外膜电位中推断这些电活动是可能的,因此,一种从体表数据恢复心外膜电位的无创方法在实验和临床研究中很有用。在本研究中,一种使用蒂霍诺夫正则化的逆方法被证明能够以良好的精度重建心脏兴奋中的重要事件。该逆过程应用于从人体躯干模型水槽中获得的数据,在该水槽中,一颗跳动的犬心被放置在正确的解剖位置。与实际测量的心外膜电位进行比较表明,在整个QRS波期间,电位特征(最大值、最小值、零电位线、鞍点等)的位置和形状都能以良好的精度恢复。位置误差通常高达1厘米,而幅度略有减小。此外,应用范围从上述几何形状精确已知且大量导联电位测量准确的设置扩展到更具临床和实验代表性的情况。研究了心脏位置不准确的影响。设计了一个近似实际几何形状的简化躯干模型,并评估了其在逆计算中的性能。提出了一种系统减少体表导联数量的方法,并根据逆解的准确性对所得的导联配置进行了评估。结果表明,逆问题可以针对测量数据中不同类型的不确定性进行稳定处理,并为在临床和实验情况下使用逆过程提供了前景。