AN EVALUATION OF RADIOLOGICAL METHODS FOR THE DETERMINATION OF LUNG VOLUMES.

Abstract

Our interest in roentgen estimates of lung compartments developed indirectly from a program evaluating the respiratory status of parents of children with cystic fibrosis. Multiple pulmonary function study technics were used, supplemented by chest films obtained on inspiration and expiration (1, 2) and roentgen kymograms (3). Equal numbers of controls were studied to make a total of 250 subjects. The purpose of this investigation was to determine whether simple radiologic measurements may be utilized to calculate the intrathoracic gas volume in a manner sufficiently accurate to be of practical research and clinical value. In 1918 Lundsgaard and Van Slyke reported a correlation of the external measurements of the chest with determination of lung volumes by gas dilution methods (4). Binger and Brow in 1924 (5) noted poor correlation of external measurements with “functional residual air” and suggested an estimate be made from chest radiographs which revealed more about the diaphragmatic level and corresponding lung volume. In 1933 Hurtado and Fray (6) were the first to actually correlate lung volumes with radiological measurements in 50 normal males. They double-exposed a postero-anterior chest film on inspiration and expiration, measured areas with a planimeter, and developed regression equations to determine the lung volumes and compartments. In correlating gas dilution lung volume with the xray method and with the external body measurement method of Lundsgaard and Van Slyke, they found the best values were obtained with the radiologic chest volume (ReV) (r = 0.636 for RCV and total pulmonary capacity, r = 0.717 for RCV and vital capacity). Wade and Gilson (7) compared radiologic chest volume determinations with lung volumes measured by conventional means. They emphasized that spirometric control of maximal inspiration and expiration is essential for good correlation (the value of r = 0.82 for total lung volume and RCV in 25 normal males with spirometric control dropped to r = 0.47 in a group of 38 normal males with no spirometric control). Cobb et al. (8), utilizing routine inspiration posteroanterior and lateral chest films from hospital files of 76 patients, measured areas from these films and introduced the figures into the Hurtado and Fray regression equations to calculate a total lung volume. They also subtracted volumes of diseased lung, yielding a corrected RCV with a better correlation coefficient (but not statistically significant) as compared to a correction for heart and mediastinal contents. The best coefficient of correlation was obtained in patients without disease masses (r = 0.906). A preliminary evaluation of 58 consecutive subjects who have been through our program revealed encouraging correlations with physiologically determined volumes as shown in Figures 1 to 3 (r = 0.880 for inspiration ReV and total lung capacity).