338 



HANDBOOK OF PHYSIOLOGY 



CIRCULATION I 



of several dipoles. It is evident, therefore, that both 

 methods of obtaining "correct" leads require simpli- 

 fications with resulting limitation of their validity. 

 Thus the image surface seems to be extremely sensi- 

 tive to changes in the location of the dipole ( 1 97, 200), 

 which has to be chosen more or less arbitrarily. The 

 anatomical data of the heart are fully neglected. The 

 lead fields, on the other hand, are constructed on the 

 theoretical basis of simple hydraulic or electric 

 models. The enormous complexity of the real situation 

 in the thorax is never taken into account. Models, 

 either of the lead field or the image surface and lead 

 vectors, do not provide a solution with general 

 validity; they remain applicable only for that indi- 

 vidual thorax form and heart position from which 

 they are taken. If, therefore, a corrected lead system 

 is to be utilized to record the EC'.G of a given person, 

 a highly complicated mathematical and experimental 

 procedure has to be put into operation, which by 

 itself excludes a broad clinical application. Neverthe- 

 less, a practical application could be to use only such 

 electrode systems as have proved, in such experi- 

 mental research, to have a statistical minimum of 

 deviation from the "best" method of recording the 

 electrical events of the heart. The question remains, 

 however, what the "best" method is and what kind 

 of electrical events the investigator wishes to receive. 

 The best method depends upon the purpose of the 

 record. There are two such purposes: the investigation 

 of the resultant heart vector as the equivalent repre- 

 sentation of all active fibers, and the derivation of 

 local events by their proximity potentials. It is 

 obvious that proximity potentials can only be picked 

 up by "precordial" leads or leads near the heart, 

 e.g., in the esophagus. The theory of such electrode 

 systems will be given in the next section. If, however, 

 the resultant heart vector is to be recorded, the in- 

 fluence of proximity potentials should be eliminated. 

 This is the case if, and only if, the lead field penetrates 

 the heart in parallel and uniform flow lines. The 

 electrode combination is best, which most nearly 

 realizes this condition. If possible, therefore, we 

 should test all electrode systems with respect to their 

 lead fields. 



6. DIFFERENT LEAD SYSTEMS 



It is not the aim in this review to enumerate or 

 even discuss the details of electrocardiographic 

 practice. For one thing, theoretical reasons for choice 

 of a certain lead position are usually not available. 



EINTHOVEN 



BURGER 



FIG. 18 Comparison between an Einthovcn and a Burger 

 triangle of the same subject. .-1: the projection of a given heart 

 vector HV on the three leads is I, II, III. These projections 

 do not represent the actual ratio of the amplitudes recorded 

 with the leads. These are correctly given in B where the pro- 

 jections I, II, and III are drawn on the sides of a "distorted" 

 triangle. The Burger triangle may be electronically reduced to 

 a correctly equilateral triangle by subdividing the lead con- 

 nections by resistances. The corrections in B are based on torso 

 models, which do not take into account the inhomogeneities 

 of the field. 



Decision as to which of the different lead systems may 

 be the best cannot be drawn out of a comparison of 

 their results, as it is unknown whether several systems 

 which yield equal results re\eal a common error or a 

 common truth. The question therefore depends merely 

 on the result of model experiments. The models 

 used so far, however, are rather simplified hoinogene- 

 ous torsos which disregard individual deviations from 

 thorax to thorax, the large differences in conductivity 

 of lung and mediastinum, and the fact that tlie normal 

 EGG is led off by flow lines which apparently do 

 not penetrate the heart-lung boundary. Nevertheless, 

 these model experiments are, for the time being, the 

 only way to compare the correctness of leads. 



A rather comprehensive work comparing nearly 

 all lead positions has been done by Schmitt & 

 Simonson (428, 429, 431). In these papers, the lead 

 vectors for the different electrode combinations are 

 given. Before we discuss these results, we should 

 briefly mention what principal differences exist in 

 the various leads and how they can be put into a 

 general order. 



Regarding technique, we may distinguish between 

 unipolar, bipolar, and multipolar leads. Each of these 

 lead systems may be used in two ways. First, one 

 could try to determine the position and magnitude 

 of a "heart vector," representing all electric .sources 

 of the heart with the same weight. We will call such 

 systems "total leads" or "heart vector leads" in the 

 following lines. The second aim of both unipolar and 



