Clinical feasibility of using an electronic portal imaging device for position verification during conventional radiotherapy

The aims of this study were to evaluate the clinical feasibility of using an Electronic Portal Imaging Device (EPID) for patient position verification, and to analyze for the EPID's limiting factor in the use of patient setup verification. The investigation involves 15 patients with different carcinomas. All of them were initially positioned by using a three-point laser alignment method and received position verification with an EPID. In the following treatment for each patient, the radiation physicians and therapists used an EPID for portal imaging to correct patient setup errors, and simultaneously used an optical patient positioning system to detect the target center positions before and after an EPID correction. By subtracting the position before correction from the after one, we obtained the data of the internal target position corrections. We acquired 44 positions denoted as three elements in left-right (LR), superior-inferior (SI) and anterior-posterior (AP) directions. 70% of the LR and SI positions lay in the range of [-4mm 4mm], and 45% of the AP positions lay in the same range. While the slight differences within this range can be hardly detected by daily EPIDs. This study indicated that the patient setup corrections according to the EPID images went largely beyond the EPID imaging resolution. And the sagging effect from gantry rotation increased the inaccuracy of patient setup verification. It appeared that the image quality and the sagging shift were the major limiting factors for the use of EPID in precise patient setup verification.