Experimental Study Of Two-parameter Exponential Model For Dosimetric Verification Of Intensity Modulated Radiotherapy

Radiotherapy, surgery and chemotherapy compose of three major treatments of tumors. Domestic and international statistics show that about 70% of cancer patients need to receive radiotherapy at different levels (radiotherapy alone or with surgery, chemotherapy). The fundamental aim of radiotherapy is to improve local tumor therapeutic gain ratio, that is, a higher tumor control probability without causing unwanted radiation damage to normal-tissues (i.e. a low normal-tissue complication probability). In recent years, the development of radiotherapy techniques is very fast. The clinical application of IMRT is a historic advancement for radiation therapy. The shape of high iso-dose surface is agree with that of target in three dimension, and the dose gradient is steeper. It has accurate target dose, small dose of the adjacent normal organ, protecting critical organ and homogeneous dose distribution. However, the process of IMRT is complex. For example, the error in tumor defined of CT, identification of the target volumes and organs at risk and patient position will be large difference of dose distribution. ICRU Report No.24 recommends a target dose uniformity lower than±5%.the complexity of the treatment may increase the probability of errors in the treatment process. And the error will seriously affect the treatment efficacy, which means it may reduce the tumor control rate and increase the normal tissue complication. Therefore, the QA and QC of IMRT become the urgent problems to be solved and should be pay great attention. The dosimetric verification is extremely important.For conventional therapy or conformal radiotherapy, we can check the dose of one or a few points to verify the plan and treatment delivery by manual. But the field shapes of IMRT are defined by a multileaf collimator (MLC). Every field contains a number of segments. The shapes and MU of these segments are different, so the dose distribution, composed by segments, can not be obtained. But for complex radiotherapy techniques, it is not sufficient because the tumorcidal high-dose volume conforms to planning target volume (the heterogeneous portal fluence and steep dose gradients).The film dosimeter, thermoluminescence gosimeter, gel dosimeter are effective tools for verification, but rather time-consuming and labor-consuming for use. As the advanced two-dimensional real-time dosimetric verification systems, two-dimensional ionization chamber matrix and EPID can obtain the two-dimensional data quickly. According to these factors, reference dose for each IMRT plan must be measured. The scatter contributions from distant beam segments and transmission through MLC leaves are significant for IMRT. To this problem, previous approaches have a modified Clarkson integration technique (using tissue-maximum ratios extrapolated to zero area to achieve scatter contributions) and convolution calculations based on an analytical pencil beam kernel.Absolute dose verification and relative dose verification are two aspects of dosimetric verification. Absolute dose verification is the way to check whether the dose at one point calculated by the quality assurance plan, which is produced by loading the IMRT treatment planning into the phantom, consists with the corresponding measurement. Relative dose verification is the way to check whether the referenced dose distribution on one plane calculated by the quality assurance plan consists with the corresponding measurements.This paper describes a simple model which separates primary and scatter dose contribution, so analytical expression of differential scatter can be received. Dose to a reference point for an arbitrary number of irregular is calculated by summing individually calculated contributions. This model is based on easily measured beam data such as percentage depth-dose and relative output factors. This provides a secondary check on more complex treatment planning algorithms. Through experiment this model is shown to provide good agreement with treatment planning system predictions. On some degree it avoids dose calculation before each IMRT planning, so significantly reducing the time required for IMRT QC.The measured data such as PDD and relative output factor come from SIMEMS PRIMUS 6MV X-ray. All plan data such as SSD and monitor units (MU) and MLC and back-up jaw positions for each beam segment are read from beam configuration files produced by treatment planning system.Implement of two parameter exponential model:first, the model was obtained through the procedure; second, plan data such as MLC and jaw position are read from TPS to determine the aperture of each segment. The parameter a and b were gotten by fitting the function; After position fixed, made the anchor, The patient data was obtained by the CT simulator. The target volume and organs at risk limits were got by CADPLAN (helios) IMRT plans., Phantom do using the same conditions and CT scan; the fluence map and the angle of gantry and the bed parameters received by IMRT reverse treatment plan, planted in the phantom imaging reconstructed by CT, the dose distribution in phantom was gotten. The dose of reference point measured by the ionization chamber were received; the position of MLC leaves and the shape of the field were read from the configuration files, so the dose distribution of the primary and scatter ray were calculated using in the center axis of different depths; according to the requirements of IMRT, the phantom was precisely positioned, and then the IMRT plans were implemented. The dose in the center axis was measured by the ionization chamber, which was compared with the predicted dose by the model and treatment planning system prediction.Results:the rage difference between the model's prediction and treatment planning system prediction was found to be±3%. The model is shown to a reliable and accurate independent check of planning system monitor units for IMRT plans. The head scatter and phantom scatter contribution was not differentiated. The transmission through MLC was not taken into account.The end part of the paper gives an overview and puts forward the issues to do in the future.