Measurement of the Heart Size with the Roentgenkymograph

The estimation of heart size in the living subject by the usual X-ray methods is generally inadequate in one or more respects: (1) the phase of the heart cycle is neglected—-the total stroke output of the heart varies from 6% to 40% of the systolic heart volume, (2) it is assumed that simple linear measurements (e.g., transverse diameter) accurately reflect true heart size (volume), (3) proper correction is not made for body size, (4) correction for distortion in the teleroentgenogram is disregarded, (5) accuracy is not proved by controls and duplicate measurement, (6) pendulum motion of the heart is included.

The roentgenkymograph, in which the film is moved during exposure behind a lead grid (Roesler; 1 Stumpf, Weber and Weltz; 2 Keys and Friedell 3 ) registers the systolic and diastolic excursions of the heart contour, eliminates the effect of pendulum motion, and gives better visualization of the upper and lower borders.

We consider heart volume to be the ultimate reference measurement. Except with cadavers we have used frontal roentgenkymographs made at 66 to 72 inches with 2.5 second exposure, and have obtained the correction factors for distortion from lateral plates. All measurements were made on tracings of the X-ray plates and independently checked. Areas were measured with the planimeter.

Measurement of the frontal systolic silhouette area is very precise. Kymographs were repeated on 8 subjects at intervals of 2 to 10 minutes. The average deviation from the means for the 8 subjects was ±0.55% and the maximum deviation was only 1.6%. Three other subjects were studied from 4 to 6 times each over a period of 3 months. The average deviations from the means of the systolic areas for these subjects were ±0.7%, ±1.9%, and ±1.1%.