A comparison of air‐cavity inhomogeneity effects for cobalt‐60, 6‐, and 10‐MV x‐ray beams

The inclusion of air-filled spaces in treatment fields creates a potential dosimetric problem due to the loss of charged particle equilibrium near the air-tissue interface. We have used a simulated larynx phantom and a small buildup/extrapolation chamber to compare the magnitude and spatial extent of underdosing and overdosing at the distal surface for two linear accelerators (10- and 6-MV x-rays) and a cobalt-60 machine. Surface doses were compared to doses measured in a similar but homogeneous phantom to give observed/expected ratios (O/E), which were greater than 1.0 for large field sizes and less than 1.0 for small field sizes on all machines. The minimum field sizes which produce no surface underdosing for a simulated 2-cm-diam larynx are roughly 7 X 7 cm for 10-MV x-rays, 6 X 6 cm for 6-MV x-rays, and 5 X 5 cm for cobalt-60. In addition, the depth over which underdosing occurs is seen to increase with increasing energy.