Since the discovery of X-rays in 1895 and its therapeutic effects shortly after, radiation therapy has continually evolved over the past century to become one of the leading treatment methods for cancer patients.
However, the effective modalities employed today—such as intensity modulated arc therapy (IMRT) and volumetric modulate arc therapy (VMAT)—come at the cost of increased sophistication in treatment planning. As further advancements are made to radiotherapy treatments, we are reaching a turning point where it is no longer intuitive for the human planner to design treatments.
The process is increasingly more time consuming as the human has to make more trial-and-error guesses in planning parameters in order to achieve an acceptable dosimetry and deliverability. Attempts to automate the processes through optimization techniques have alleviated some of the planning complexity.
However, it is difficult to manually quantify every aspect of the patient that leads into the physician’s prescription decision, resulting in multiple iterations of planning by the human planner in order to meet the physician’s acceptability criteria.
Recently, artificial intelligence (AI) has made quantum-leap advances in various areas. The substantial progress in AI development is attributed to the self-learning capabilities of recent deep learning systems. We realized that AI techniques may also be applied to solve the treatment planning problem in radiotherapy.
We aim to design an AI treatment planner capable of learning radiotherapy planning, improving itself by continually developing new plans, and eventually outperforming human planners while maintaining individual physician preferences.
Currently, we are working on developing AI that can learn the correct beam angles for treatment planning to test the feasibility of learning other planning parameters for radiotherapy. Furthermore, we are investigating the viability of generating an optimal dose distribution for radiotherapy using deep learning.
- Nguyen, D., Long, T., Jia, X., Lu, W., Gu, X., Iqbal, Z., & Jiang, S. (2017). Dose Prediction with U-net: A Feasibility Study for Predicting Dose Distributions from Contours using Deep Learning on Prostate IMRT Patients. arXiv preprint arXiv:1709.09233. (arXiv)
- Nguyen, D., Jia, X., Sher, D., Lin, M. H., Iqbal, Z., Liu, H., & Jiang, S. (2018). Three-Dimensional Radiotherapy Dose Prediction on Head and Neck Cancer Patients with a Hierarchically Densely Connected U-net Deep Learning Architecture. arXiv preprint arXiv:1805.10397. (arXiv)