Browsing by Subject "Intensity Modulated Radiation Therapy"

Translational cancer research has been around for many years and has resulted

in many advancements in cancer treatment. Preclinical radiation therapy is an important

tool used in some studies to better understand the biological effects due to radiation.

Current preclinical radiation treatment techniques do not emulate the advanced

techniques used in cancer clinics, such as intensity modulated radiation therapy (IMRT).

In this work we explore the possibility of developing and implementing an IMRT

treatment capability for an orthovoltage micro irradiator used for small animal research.

In order to implement IMRT to the micro irradiator, every step of the radiation

therapy treatment process had to be evaluated, developed, and tested. The first step was

to develop and treatment planning software that can be used for small animal studies.

Using the open source Computational Environment for Radiotherapy Research (CERR)

and adapting it for use with an orthovoltage irradiator, monte carlo dose calculations

could be performed for small animal data sets. CERR does not have the ability to

optimize dose calculations, so a Matlab script was developed and written for inverse

optimization for treatment planning. Treatment plans were designed and optimized for

several small animal cases to evaluate the optimization algorithm. Following successful

simulation development, treatment delivery techniques needed to be developed. 3D

printing was used as a tool to create physical compensators that could be used as an

add-on device to the micro irradiator. With the capability of submillimeter printing

resolution, 3D printing has the capability to handle the high resolution required for very

small structures inside of small animals. Using the simulation data, another Matlab

script was developed to create both compensator and inverse compensator 3D models.

Many materials and techniques were evaluated to determine the best method for

compensator production. Materials were tested for attenuation properties, printing

capabilities, and ease of use until a satisfactory result was achieved.

Once the simulation and delivery techniques were developed to a satisfactory

level, an end to end test was designed to verify the IMRT capability. Using a 2.2 cm

diameter cylindrical Presage® dosimeter as the quality assurance (QA) device/patient, a

treatment plan was created based on the geometry of the Radiologic Physics Center

(RPC) Head and Neck phantom design. The dose tolerances used for the inverse

optimization were the same as the RPC Head and Neck protocol with a stricter tolerance

for the organ at risk (OAR). Compensators were produced for the plan and both 2D and

3D analysis was performed. Radiochromic film was used for 2D dose map analysis.

Gamma analysis was performed using 2D film data with varying criteria for distance to

agreement and dose difference. 3D analysis was done by delivering the treatment plan

to the Presage® dosimeter. Using optical-CT for dose readout of the dosimeter,

qualitative analysis was performed to show the 3D delivered dose data.

The end to end test showed strong evidence that IMRT could be implemented on

the small animal irradiator. The 9 field treatment plan was delivered in under 30

minutes with no mechanical or collisional issues. The 2D dose analysis showed 7 out of 9

treatment fields had a passing rate greater than 90% for a gamma analysis using 10%/0.5

mm tolerances. 3D dose analysis showed promising spatial resolution of the dose

modulation. As a feasibility and an initial testing study for a new treatment technique on

the small animal irradiator, these results showed the capability of the 3D printed

compensators to modulate dose with high spatial precision and moderately accurate

dose delivery.

Introduction: The current head-and-neck (HN) fluence map generator tends to producehighly modulated fluence maps and therefore high monitor units (MUs) for each beam, which leads to more delivery uncertainty and leakage dose. This project implements diffu- sion into the training process and modifies the loss functions to mitigate this effect.

Methods: The dataset consists of 200 head-and-neck (HN) patients receiving intensity mod-ulated radiation therapy (IMRT) for training, 16 for validation, and 15 for testing. Two models were trained, one with-diffusion and one without. The original model was a con- ditional generative adversarial network (GAN) written in TensorFlow, the model without diffusion was written to be the PyTorch equivalent of the original model. After confirming the model was properly converted to PyTorch by comparing outputs, both new models were modified to use binary cross entropy for the GAN loss and mean absolute error as a third loss function for the generator. Hyperparameters were carefully selected based on the training script for the original model, and further tuned with trial and error. The diffusion was implemented based on Diffusion-GAN and the associated GitHub repository. The two new models were compared by plotting training loss vs epoch over 500 epochs. The two models were compared to the original model by comparing the output fluence maps to the ground truth using similarity index and comparing DVH statistics among the three models.

Results: The with-diffusion model and no-diffusion model achieved similar training loss.The diffusion model and no-diffusion model consistently delivered better parotid sparing than the original model and delivered less dose to four of the six tested OAR. The with- diffusion model delivered less dose to five of the six tested OAR. The diffusion model had the least MUs: 23% less than the original model and 3% less than the no-diffusion model. The diffusion model had lower D2cc: 4% less than the original model and 1% less than the no-diffusion model on average. All three plans deliver 95% of the prescription dose to nearly the same percentage of PTV volume.

Conclusion: Implementing diffusion does not provide a significant impact on training timeand training loss. However, it does enable comparable dose performance to both the no- diffusion and original models, while significantly reducing the total MU’s and 3D max 2cc relative to the original model and slightly reducing these metrics relative to the no-diffusion model, indicating smoother fluence modulation. In addition, both new models reduced dose to the right and left parotids relative to the original model, and to four of six tested OAR total, while the with-diffusion model consistently delivers less dose to OAR than the no- diffusion model. This indicates that both the new loss functions and diffusion reduce the overall dose to the OARs while preserving dose conformity around the target.