Correction of lateral artifacts in table scanner algorithm for evaluation of radiochromic films

Radiochromic films are indicated to evaluate the difference between the planned dose and the dose deposited in radiotherapy. The accuracy of the dosimetry with the use of films causes a relationship between the R, G, and B levels and the radiation doses, but this relation varies with the position of the film along the scanner. To correct this problem, this study aimed to implement a free code algorithm in Python, using the methodology proposed by Lewis and Chan (2015). For this, it was necessary to raise by an empirical process, the relationship between the levels R, G, and B (red, green and blue) and doses. After surveying mapping functions between levels and doses, a scanned image could be converted into a dose image. Thus, the films were prepared, irradiated and scanned at determined positions at time intervals. Scanned images were separated at R, G, and B levels and analyzed with ImageJ. The values of dose levels were used in the implementation of algorithm codes contemplating the channels and the positions of the film. The levels of R, G, and B were corrected in the original images by the algorithm using linear interpolation, to generate new processed image for each position as a function of the levels of the center of the scanner. The algorithm was validated and approved with the correction of all regions of movie artifacts. An example is the average of the two extreme positions of the red channel, which the deviation of the original image was 5.8%, while the deviation of the processed image was 1.6%. The influence of the time effects between the irradiation and the scanning of films with 1% deviation for the period from 12 to 72 hours was increasing. The algorithm equivalently corrected the processed images even taking into account only the position in the correction process. The differences between the levels of R, G, and B and the optical densities of the original and processed images were similar. With this study, it is hoped to extend the application of the algorithm to all radiotherapy treatment sites. (author)