Development of a Method for Automated Estimation of Tumor Location on EPID Cine Images in Lung Stereotactic Radiotherapy

Abstract

The purpose of this study was to develop a method for automated estimation of the tumor location in cine images on an electronic portal imaging device (EPID) without using implanted markers during lung stereotactic radiotherapy (SRT). First, the irradiation field images were determined by analyzing the histogram and cropped from all EPID cine images. Second, a tumor region in the reference image of the EPID cine images was enhanced by using a Gaussian image enhancement filter (GIEF), and then it was used as a “tumor template” image segmented based on a multiple-gray level thresholding technique and a region growing technique. Finally, the tumor location was determined by the position in which the tumor template image took the maximum cross-correlation value within each irradiation field image following the reference image. EPID images with 512×384 pixels (pixel size: 0.56 mm) were acquired at energies of 4 or 6 MV on linear accelerators. The proposed method was applied to EPID cine images (228 frames) of 13 clinical cases with a non-small cell lung cancer. As a result, the average location error was 1.63±0.62 mm. This examination result suggested that physiological motion of tumor can be monitored by using the proposed method during radiotherapy without implanted markers.