Abstract
The sensitivity of EPID-based transit dosimetry to detect patient variations between treatment fractions is examined using gamma analysis and a structural similarity (SSIM) index. EPID images were acquired for 3-dimensional conformal (3DCRT) and dynamic intensity modulated (dIMRT) radiation therapy fields in multiple fractions. Transit images were converted to doses, transit dose in the first fraction considered the reference dose. Variations in patient position or weight were then introduced in the subsequent fractions. Positional variations were examined using a lung and a head and neck phantoms. Anatomical variations were examined using a slab phantom in three scenarios, with solid water simulating tissue, medium-density fiberboard simulating fat, and Styrofoam simulating lung. The dose difference between the first and subsequent fractions was computed using various gamma criteria and the SSIM index. Using a criterion of 3%/3 mm, EPID can detect positional variations >= 4 mm, and tissue and fat variations >= 1 cm, whereas it cannot detect lung variations up to 4 cm. The sensitivity for 3DCRT is higher than for dIMRT. EPID can detect the most variations when using 3%/1 mm. With the SSIM index, EPID can detect a 2 mm positional variation and 1 cm of lung variation. The factor that optimized the sensitivity of EPID was a reduction in the distance to the agreement criteria. Our study introduces the SSIM as an alternative analysis with high sensitivity for minimal variations.