FIGHT CANCER WITH CODE, JOIN US!
Physics Master Thesis: Improving an analytical dose computation for proton radiation therapy
We offer three physics master thesis positions at RaySearch Laboratories in Stockholm. Six topics are formulated and this announcement describes one of them.
The topic is to investigate a set of new ideas for analytical proton dose calculations. By combining two existing algorithms, we believe that substantial accuracy improvements for lung tumors and areas close to bone could be achieved. Moreover, the current solution for handling protons from nuclear scattering leave room for improvement and creativity.
Your tools for this task are RaySearch’s powerful treatment planning software RayStation, the RayStation scripting interface in Python and the RayStation development code in C++ and C#. As support you will have your supervisor at RaySearch, the RaySearch crew in general and the physics group at the Development Department particularly.
Background and purpose
A treatment planning system (TPS) is a highly advanced software package that is used in the radiation oncology clinic to generate radiation plans for cancer patients. The overall objective of the TPS is to create treatment plans that give the prescribed amount of absorbed radiation dose to the tumor, while at the same time spare the surrounding healthy tissue as much as possible. One of the most delicate components of any TPS is the dose engine, i.e. the algorithm used to predict the absorbed dose in the patient. A treatment plan is never more successful than the accuracy of the calculated dose. An error in the dose computation can lead to the creation of sub-optimal treatment plans and potentially mistreatment.
Proton radiation therapy is a treatment technique that has more-or-less exploded in popularity the last 5 years. It is a far more expensive technique than conventional radiation therapy, but offers substantial advantages in terms of healthy tissue sparing. Traditionally, the dose engine used for proton radiation in commercial TPSs are based on analytical algorithms. There is a great variety of these algorithms, but what they all have in common is that they transport phase space distributions of sub proton beams through the patient, and scale a reference depth dose curve to each point in the patient, taking the inhomogeneities in the patient elemental composition into account. These algorithms have for the most part been successful in predicting the dose in the patient, and can be highly computational efficient. However, there are situations where the analytical proton algorithms are less successful. These situations include treatment sites that host large lateral inhomogeneities along the beam path. Examples include treatment of lung tumors. Another situation that has been difficult for the analytical proton dose engines are setups that include a so-called range shifter in the beam, a commonly used device for reaching shallow tumors with the protons. The problem here is that the analytical dose engine does not handle transport of secondary protons generated in the range shifter particularly well.
Just recently, the two dominating TPSs for proton therapy (RayStation from RaySearch Laboratories, and Eclipse from Varian) have released proton dose engines based on the Monte-Carlo (MC) principle. The MC based dose engines solve the above-mentioned problems, but it comes at a price, namely calculation time. Dose computation times may not be critical for the creation of single plans, but becomes limiting for more advanced planning and plan evaluation workflows, where several hundred dose calculations for the same patient may be required. This means that the need for fast, reliable analytical proton dose calculations is still of essence.
This master thesis project aims at investigating some new ideas for analytical proton dose calculations. By combining two existing algorithms dealing with lateral inhomogeneities, substantial improvements for e.g. the lung case could be achieved. Moreover, the current ad-hoc solutions for handling secondary protons in the analytical algorithms leave a lot of room for improvements. The work will mainly be conducted with the current analytical dose engine in RayStation as starting point, with the ultimate goal to implement the new algorithms in a subsequent release of RayStation (not part of the Master thesis project).
Each improvement can be evaluated against the RayStation Monte Carlo, clinic measurements and independent Monte Carlo computations, using real patient plans and geometries or idealized cases which provide physics insight.
Education, experience and knowledge
This master thesis is the final part of a master education in physics, engineering physics, medical physics or a similar program, where you have excelled.
You are motivated and intelligent and you like physics analysis and programming. You are used to computers and programming, if the experience is less this is compensated by will and skill to learn.
Meriting but by no means required are proven skills in C++, less likely but still meriting is knowledge of C#/.NET.
Spoken and written English is required, Swedish is highly meriting.
You have good analytical skills and a problem-solving mindset. When working on a problem, you formulate the right questions, identify what is important and work in a goal-oriented way towards a solution. You are a strong team player who takes responsibility and contributes to a positive atmosphere. You take pride in delivering high-quality results with efficiency.
About physicists at RaySearch
The physicists at RaySearch are responsible for physics algorithms and functionality for photon, electron, proton and carbon-ion treatments within the treatment planning system RayStation. The physicists participate in all steps of the research and/or development processes, including method investigations, data analysis, software design, implementation and testing, at times in collaboration with external partners or clinics.
Working at RaySearch
RaySearch believes in investing in its people. We prioritize knowledge-sharing, creativity and collaboration, and you will work together with some of the most talented and highly educated people in the industry. We also have a strong social culture, with regular events and activities for employees. You will work in a modern office environment, with access to the latest hardware and tools. RaySearch is committed to equal opportunities. We value diversity and are dedicated to preventing discrimination. Read more about RaySearch.
You are welcome to send your application (preferably in Swedish but English is also fine) by clicking on the button below. It should include a résumé, a personal letter and documentation of your university education (BSc and MSc). Mark the application “Physics Master Thesis 2018”.
Three positions are announced, but six topics are formulated. The recruitment process for the three positions will go through the same contact point. Please send in one application and indicate clearly all projects of interest (including your thoughts on what would suit you and interest you the most).
The recruitment process will be ongoing while this posting is present. To enquire about the status of your application, please email firstname.lastname@example.org.