Table of Contents Table of Contents
Previous Page  1018 / 1096 Next Page
Information
Show Menu
Previous Page 1018 / 1096 Next Page
Page Background

S1002

ESTRO 36

_______________________________________________________________________________________________

The plans were then recalculated implementing the shifts

using the algorithm used for the clinical plans (Eclipse ™,

Varian Medical Systems, Palo Alto, AAA algorithm, v 13.6).

The mean and maximum doses for the lungs, kidneys,

brain and the (body-lungs-5mm) structure were extracted

and the difference between the planned and the

recalculated

doses

determined.Results

The mean doses change by a maximum of 0.6% (lungs), 0.6

(kidneys), 0.5% (brain) and 0.2% (body-lungs-5mm). The

greatest difference between the maximum doses are 8.0%

(lungs), 4.8% (kidneys), 2.6% (brain) and 12.0% (bodylungs-

5mm).

The standard deviation of the difference between the

calculated and recalculated doses are greater for the

maximum doses than the mean doses (figure 2). Given that

the minimum and maximum doses for SS TBI are typically

in the range 90-110% of the prescribed dose, the

differences in maximum dose should lead to care

being

taken when positioning patients for SS TBI.

Conclusion

Patient positioning for a total of 63 fractions of SS TBI is

such that the mean delivered doses differs from the

planned by less than 0.6%. However, the maximum doses

are more sensitive to incorrect patient positioning,

differing by up to 12% with the delivered dose being

greater than the maximum. Correct patient positioning or

SS TBI is pertinent.

EP-1830 Simple method on bladder filling simulation to

improve the soft-tissue evaluation on CBCT

K.L. Jakobsen

1

, K. Andersen

1

, D. Elezaj

1

, D. Sjöstrøm

1

1

University Hospital Herlev, Department of Oncology,

Herlev, Denmark

Purpose or Objective

The purpose of this study is to present a cost effective

method on how to evaluate the robustness of the

treatment plan on different bladder fillings during

treatment planning. Furthermore the purpose is to

evaluate how this method can be used to determine when

a bladder is too small during treatment of the patient.

Material and Methods

Patients suffering from anal and rectum cancer were

enrolled in the study. All patients were instructed to

follow our bladder protocol where the patients are asked

to empty their bladder 1 hour prior to scan/treatment and

then drink 2 glasses of water. The bladder and the bowel

were delineated on the CT image set according to

QUANTEC guidelines. At the treatment planning stage

different bladder fillings were simulated by cutting off ¼,

½ and ¾ of the bladder in the cranial-caudal direction

(Figure 1). By using the different bladder volumes the

corresponding bowel volumes were created. The

robustness of the treatment plans was evaluated by

identifying if the bowel constraint was fulfilled for the

different simulated bladder fillings. If bowel constraint

wasn’t fulfilled the treatment plan was re-optimized to

improve the robustness. Before each treatment CBCT was

acquired and the true bladder filling was compared to the

simulated situations. For the situations where the bladder

filling was identified to be too small so the bowel

constraint was violated the patients were asked to drink

more water. For some of the patients the true bladder was

delineated on CBCT and the corresponding bowel was

generated and compared to the simulated situation.

Results

For most of the rectum cancer patients the constraints was

fulfilled for all simulated situations. Due to the higher

prescription dose and also the location of the target the

anal cancer patients didn’t match the constraints to the

same extent. The study revealed that most of the

treatment plans was robust to bladder filling changes but

also identified situations were re-optimization could be

done to create a more robust treatment plan (Figure 2).

The RTTs found it feasible to compare the bladder on the

CBCT with the simulations

and was also able to identify

when additional actions were needed.

Conclusion

This procedure has shown to be very cost effective as it

doesn’t require additional imaging and it only takes 10-15

minutes to create the simulated structures. The latter can

be optimized further in the future e.g. we consider to only

simulating the smallest bladder (largest bowel) for the

rectum cancer patients. This should be compared with our

previous workflow with unreasonable demands on bladder

filling and delineation of the bladder on CBCT with the

rather subjective decision when the bladder was

considered to be too small. Furthermore this workflow has

made it able for the RTTs to get more involved in

evaluating and react on differences in soft tissue.

EP-1831 Entropic Boltzmann closure for MRI-guided

radiotherapy

J. Page

1

, J.L. Feugeas

1

, G. Birindelli

1

, J. Caron

1

, B.

Dubroca

1

, T. Pichard

1

, V. Tikhonchuk

1

, P. Nicolaï

1

1

CELIA, Interaction- Fusion par Confinement Inertiel-

Astrophysique, Talence, France

Purpose or Objective

The majority of patients affected by cancer are nowadays

treated by radiotherapy, which consists in delivering a

homogeneous dose with energetic particles. The main goal

of this technique is to target and destroy tumoral cells

without damaging the surrounding tissue. This treatment

possesses a great adaptability to the broad variety of

tumors. Therefore, a major effort was made on the last

decades to improve technologies involved in the

development and the optimization of this treatment. Our

work consists on the development and validation of a new

model designed to simulate the energy deposition of the

particles used in radiotherapy (electrons, photons and

protons), within human tissues.

Material and Methods

This model is based on a kinetic entropic closure of the

linearized Boltzmann equation, which describes the

transport of energetic particles in the matter. This

equation takes a lot of computation time to be resolved

due to the high number of variables. To simplify this, we

replace fluences by angular moments, which allows us

getting rid of the angular variables andimprove the