"Outcome Prediction and Decision-making in Radiation Treatment Planning using Deep-Learning techniques"

The use of Artificial Intelligence (AI) has been growing in the healthcare industry including Radiation oncology. The workflow of Radiotherapy planning and delivery is complex with its requirement of multiple checkpoints like immobilization, imaging, contouring, planning, quality assurance (QA) and delivery. Many efforts are being made across the globe to automate and include neural networks in these stages to have a precise and faster outcome. Automated solutions at different stages like contouring, planning, and QA analysis are available either partially or fully. However, to start with the radiotherapy process the decision-making is still done manually and is dependent on the individual practitioner.

Hence, this project focuses on developing a module using deep learning techniques to facilitate decision-making process. The data like existing contours, beam arrangement templates, dose constraints, optimization algorithm, delivery technique and dose distribution, will be used as sample data for the neural networks. The trained neural networks will be used to compare the input patient data with the existing library. Thus the best match of the template used for a similar earlier case will be identified. The developed module can help in saving lots of time for the specialists involved in making their decisions on best treatment plan, as they would not require several trials and errors.