Optimisation in radiation treatment planning

Radiation treatment (also called radiotherapy) is one of the most common treatment methods for cancer, with almost two-thirds of all cancer patients expected to have it at some stage in their treatment plan.

The treatment uses high-energy radiation to kill cancer cells by damaging their DNA. There are various ways to deliver the necessary radiation:

• radioactive sources can be placed in the body near the tumour. This is called brachytherapy.
• external radiation therapy, where radiation beams produced by a Linear Accelerator (LINAC) are aimed at the tumour and its surrounding tissues to target cancer cells and their potential spread.

One of the key challenges in radiotherapy is the treatment planning. Each individual case is unique and therefore a treatment plan specific to each patient is required, which uses detailed information generated through CT scans. The availability of radiotherapy equipment is often limited, since such equipment require substantial investments. Therefore, it's crucial to be able to create the most efficient treatment plans, both for the sake of treating each individual case and for making the best use of expensive machinery.

Dr Kerem Akartunali and his colleagues worked on developing some efficient algorithms optimising treatment plans using a variety of different external radiotherapy machinery [1]. In this work, they developed optimisation models considering various decisions such as how much radiation to beam from a specific angle and how to shape the associated pattern. In real world applications, such models can easily become computationally intractable due to huge number of decisions variables to consider. Therefore, they've developed some efficient heuristic algorithms that can be indeed used in practical applications, showing that some of these algorithms can be very efficient compared to other existing techniques.