Clinics store a vast amount of patient data, including a huge treatment plan database of past treatments. These plan databases can be potentially exploited by comparing the database contents to the current situation. We hypothesize that approximating the current treatment plan by finding and using the most similar one in the plan database will significantly speed up the treatment creation process, both in case of initial treatment planning and daily plan creation.
Such comparison is already possible using our open-source radiation therapy research toolkit, SlicerRT [1]. However a manual comparison workflow of one plan to another takes several minutes, and as a plan database potentially contains thousands of plans, finding the most similar plan is not feasible in the required timeframe. The process can be made faster by automation, but still, a local computer executes the individual comparisons in a linear fashion, allowing for very limited parallelization.
Building the needed computational power locally to parallelize algorithmically demanding tasks was the only available approach in the past. With the advent of cloud computing, this has changed. The cloud provides elasticity and scalability of resources through dynamic allocation, potentially exercising a multitude of virtual machine instances to focus on a single problem. Using the cloud avoids upfront infrastructure costs, as users only pay for the cloud resources they use. Hospitals having access to such a vast cloud resource instead of building the computation power themselves can potentially perform highly computationally demanding applications, such as treatment plan comparison in a plan database.
This project is an explorative and feasibility study examining the applicability of cloud resources for selecting the radiotherapy treatment plan from a large plan database that is most similar to the current situation, to facilitate
- Approximation of the initial treatment plan in order to decrease work demand from physicians, and
- Approximation of the daily adaptive treatment plans to increase overall treatment accuracy, thus improving patient life expectancy.
For background information see parent page.
[1] Pinter C, Lasso A, Wang A et al. (2012) SlicerRT: Radiation therapy research toolkit for 3D Slicer. Medical physics 39(10): 6332-6338.
