Prostate IMRT: Two-dimensional model of rectal NTCP employing the variability of rectal motion and rectum wall thickness

In order to improve the evaluation of possible rectal toxicity based on the rectal normal tissue complication probability [NTCP], we consider the fractional dependence of the NTCP on the wall thickness [t[W]] and rectal displacement [R[M]]. The two-dimensional NTCP model [NTCP[2D]] was developed using radiotherapy plans of ten randomly selected patients with prostate cancer. The clinical rectal structures were substituted with rectal walls of cylindrical shape. To simulate full, partially-full and empty state of the rectum, three tW were generated under the conditions of same length of the rectum and same volume of the rectal wall. A thrsigmaold iso-line, NTCP[TR], was used to split the NTCP[[2D]] field into areas: a lower risk area and a higher risk area for rectal toxicity. Two factors are introduced to help with the estimation of NTCP: a volume factor k[1] which is the ratio between the volumes of the rectal wall and the intersection of the rectal wall with the planning target volume; and a probability factor k[2], which is the ratio between the area of low risk to the entire area of the NTCP[[2D]]. A correlation > 0.9 between factors k[1] and k[2] was found. The NTCP[2D] field and the ratios k[1] and k[2] can be used as a patient-specific parameters to evaluate the probability of rectal toxicity

Dosimetry limitations and pre-treatment dose profile correction for sliding window IMRT

This work investigated the dosimetry limitations of the random and systematic uncertainties of sliding window [SW] intensity modulated radiation therapy [IMRT]. A Varian 21EX linear accelerator, Pinnacle[3] treatment planning system and radiographic film dosimetry was used. The limitations of the SW were studied using beam modulation ranging from 2 to 100 MU/beam, DR from 100 to 600 MU min[-1], LV from 1 to 5 cm [s-1] and field size up to 12 x 12 cm[2]. The random and systematic errors were investigated using clinical and flat beams, as well as beams of high profile modulation including linear, exponential, and sinusoidal profiles. The leading edge and plateau of the SW profiles have a significant deformation for higher DR and for beams of < 10 MUs/beam. It was found that the error is directly proportional to the DR and LV, and inversely proportional to the number of MU/beam. The high DR and LV are limiting factors, producing random profile deformation when SW beams of small number of MU/beam are delivered. A very good agreement was found between the planned and delivered geometrical and clinical dose profiles when beams > 10 MUs irradiated by a DR from 100 to 600 MU min[-1] and LV from 1 to 5 cm s[-1]. After the proposed correction, an average difference < 0.5% for clinical profiles was measured for beams irradiated with DR = 600 MU min[-1] and LV= 5 cm s[-1]. It was concluded that this correction methodology may serve as a pre-treatment Quality Assurance tool for SW IMRT beams

Prostate helical tomotherapy: a semi-empirical estimation of the scaling factor based on 2D approximating field

In Helical Tomotherapy [HT], the scaling factor [SF] is the time in seconds that each leaf viewing a target would need to be open to deliver the prescribed dose. The SF is patient-specific and is used to calculate the rotational period of the gantry, and the total treatment time [TTT] of the HT. The SF is generally difficult to estimate. Currently, it takes about one hour to fully optimize a prostate HT plan and to calculate the corresponding TTT. The aim of this study is to develop a method for estimation of the SF directly using a patient-specific approximating function. The SFs of ten randomly selected patients were used to build the approximation model. For the entire group of patients the PTV1 ranged from 57 to 396 cm3 for PTV1 margins from 2 to 10 mm. The discrete data for every patient is represented by an individual function, SF=f [kx PTV1]. The values of the function were rescaled to a special unit which represents the target volume irradiated with the prescribed dose per second. The values were normalized with two "geometric" parameters, namely, the target-to-target and the body-to-body ratios. After the normalization, the function for every patient was ordered in the file by the volume of the prostate and seminal vesicles. For prostate HT planning, it was found that the planning target volume [PTV1] has a higher impact on the SF values than the size of the patient's bodies. The function SF=f [kxPTV1] was found smooth and continuous over the given interval. The rescaled and normalized functions for all patients were represented as delimiters of a 2D field. The method proposed for determination of the SF and TTT for HT prostate planning covers PTV1 of four margins and a volume of prostate and seminal vesicles ranging from 42.8 to 161 cm3. Using these approximations, the TTTs for a second group of patients were determined in several minutes with deviation ranging from -2.8% to +7.1% compared to that of the TTTs calculated using the HT planning system