Treatment verification with megavoltage electronic portal imaging applied to the tomotherapy concept

Purpose: A new treatment strategy called Tomotherapy, introduced by T. R. Mackie et al. in 1993, was designed to perform a dynamic conformal treatment technique in precision radiotherapy. This technique delivers sliced intensity-modulated radiation fields to achieve best tumor control while sparing neighbouring sensitive normal tissue and organs at risk. The beam continuously revolves around the patient similarly to a spiral-CT while the patient is moved through the bore of the gantry. As a first step towards the realization of such a concept with a linear accelerator (Siemens Mevatron Experimental) used in clinical routine, we focused on treatment setup and dose verification. In tomotherapy, an actual CT data set is needed for patient positioning and for the verification of the absorbed dose, also, the dose transmitted through the patient must be known. This makes possible both routine tomographic treatment setup verification and tomographic dose reconstruction of the actual delivered dose. Materials and Methods: All measurements were performed with a megavoltage electronic portal imaging device of WellhoeferTM (BIS-710). The BIS-710 detector is based on a scintillation foil and contains a camera for 10-bit digital data output. The dimension of the detector plane is 512 x 512 pixels with a pixel size of 0.6 mm in each direction. The BIS-710 was developed especially for quantitative dose measuring, whereas most of the existing Portal Imaging Systems are used for image display only. To examine the properties of the BIS-710 concerning tomographical reconstruction with a therapeutic 6 MV X-ray beam, a tissue-equivalent Alderson head phantom was rotated stepwise across a stationary beam between the collimator and the detector plane. The influence of scattering can be estimated by comparing measurements which were taken with a homogeneous phantom which is invariant under rotation with a calculated exit dose distribution using a simple exponential law for the attenuation coefficient. For the calculation the attenuation coefficient distribution we adapted an iterative reconstruction algorithm. Results: Using film dosimetry, we tested the dose accuracy of the Wellhoefer imaging system. The result matches within +/- 2%. For reconstruction, we used a data set of 120 projections over the whole 360 deg. range. Each projection was measured with an array of 512 detector elements. The sample time for each image was 0.88 seconds. To perform satisfactory megavoltage CT imaging, it is necessary to sample projection data over an angular range of 180 deg. plus the fan angle. In our experimental setup the fan beam angle was about 14 deg. and the projections were taken at 2 deg. intervals. In this case, 98 projections are required and the image acquisition for all projections takes between 12 and 88 seconds depending on the resulting image resolution. This yields an additional total dose of 6 to 46 cGy per slice. Conclusion: Comparing theoretical with measured data, we observe that the megavoltage electronic portal imaging device of Wellhoefer (BIS-710) is able to give a true picture of actual exit dose distributions. Given this, we conclude that the BIS system delivers suitable data for the tomographic reconstruction of the absorbed dose. Furthermore, our first megavoltage CT image reconstruction also demonstrates the power of the BIS-710 system. However, we believe that the additional dose, given to the patient due to the megavoltage image acquisition, could be significantly reduced by operating the linear accelerator at a lower dose rate and by improving the portal image read-out software. In future investigations we plan to take into account the whole energy spectrum of the beam and perform scatter corrections to achieve even better images