Heart Measurement

Abstract

RECENTLY HODGES (1) and Schwarz (2) have described in this Journal the technical details and nomograms that have been used in this laboratory for determining heart size in the teleroentgenogram, either from planimeter measurement of frontal plane area or from linear measurement of the long and short diameters of the frontal plane silhouette. In both instances, several steps were involved in finally obtaining heart size expressed in percentage variation from normal. The purpose of this paper is to see how far one can go in simplifying heart measurement in adults without any significant loss in accuracy. It became apparent that, in order to attain this objective, three things would be desirable: (1) To eliminate calculation of divergent distortion for each individual patient and, instead, obtain an average distortion factor which could be applied to all subjects. (2) To avoid making an outline of the complete cardiac silhouette, using only those borders which can be seen on the chest film. (3) To combine in one nomogram the equation for predicting normal size with the equation for measured size in such a way that percentage variation from normal could be read directly. In order to obtain an average divergent distortion factor, one must always take the chest film in the same manner, preferably at a distance where divergent distortion is reduced to a minimum consistent with a reasonable exposure time. Since our routine chest films have been taken at a 72-inch target-film distance with the anterior chest surface closest to the film, and since this is also the common practice elsewhere, it was selected as my standard technic. The only variable then is the distance of the heart from the film, a factor which varies with the anteroposterior diameter of the chest. To determine if variation in divergent distortion between subjects was really significant, 100 adult patients were selected at random, with wide differences in height, weight, and heart size, as shown in Tables I to V. The divergent distortion correction factor was figured individually, based on the anteroposterior chest diameter of each patient, and the result was recorded. The average divergent distortion correction factor was 0.89, obtained by adding the individual correction factors together and dividing by 100. The range extended from only 0.87 to 0.91, despite the fact that anteroposterior chest diameters varied from 16.5 cm. to 29 cm. Therefore, the greatest possible deviation from the average or mean is about 2 per cent of the total frontal plane area, an insignificant figure. As the next step in simplification, one would like to avoid making a tracing of the complete cardiac outline, particularly since the upper and lower borders are drawn empirically. The product of the long and short diameters multiplied by a correction factor will give a reasonable approximation of frontal plane area.