Automated Radiotherapy Planning Across Vendors and Clinical Indications
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional radiotherapy planning is time-consuming, prone to variability, and limited by vendor-specific systems, leading to suboptimal treatment plans and inefficiencies in data exchange and interoperability.
Innovation Solution
An automated radiotherapy planning system that is indication-agnostic and vendor-agnostic, using 3D medical images and dose prescription parameters to segment target volumes and organs at risk, generate a 3D volumetric dose distribution, and optimize treatment plans with a unified platform that interfaces with multiple vendor systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual segmentation and treatment planning is performed by clinicians, then the treatment plan can be customized based on individual expertise, but the process is time-consuming and prone to variability
Solution Approach 1:
The system performs preliminary automated segmentation of target volumes and organs at risk using pre-trained deep learning models before treatment planning. This preliminary action prepares the data structure and identifies critical structures in advance, reducing the time required for manual segmentation while maintaining consistent anatomical delineation across different cases
Solution Approach 2:
The system uses pre-trained deep learning models that have been trained on large datasets of annotated medical images. These models create accurate copies or predictions of target volumes and organs at risk by learning from previously labeled data, enabling rapid and consistent segmentation without requiring manual annotation for each new patient case
2Reliability
If vendor-specific planning systems are used, then the system can be optimized for specific equipment, but interoperability and flexibility are limited
Solution Approach 1:
The system is designed with a vendor-agnostic architecture that can interface with multiple radiotherapy delivery systems from different vendors. It uses standardized data formats and protocols (such as DICOM) to communicate with various imaging modalities and treatment machines, enabling a single platform to serve multiple functions across different equipment ecosystems without requiring vendor-specific customization
Solution Approach 2:
The system acts as an intermediary layer between medical imaging systems and radiotherapy delivery systems. It receives imaging data in standardized formats, processes the information through automated segmentation and planning algorithms, and outputs treatment plans that can be delivered to various vendor-specific systems, thereby mediating between different technological platforms
3Productivity
If automated deep learning systems are used, then the planning process is accelerated, but indication-specific models limit applicability
Solution Approach 1:
The system employs a dynamic model selection mechanism that automatically adapts the deep learning model based on the specific clinical indication and anatomical region being treated. Rather than using a single static model, the system can switch between or combine multiple specialized models (e.g., for prostate cancer, breast cancer, lung cancer) depending on the input case, thereby maintaining both speed and versatility across diverse treatment scenarios
Solution Approach 2:
The system changes key parameters such as the anatomical region of interest, target structures, and dose constraints based on the clinical indication provided. By dynamically adjusting these parameters and selecting appropriate pre-trained models for each indication type, the system maintains high planning efficiency while being applicable across a wide range of cancer types and treatment scenarios
Data Source
AI summary
An automated radiotherapy planning system for generating a deliverable radiotherapy plan (31) for a patient, to be delivered by a predefined radiotherapy delivery system, the system is configured to segment a target volume and one or more organs at risks in a 3D medical image of the patient, generate a prediction of a 3D volumetric dose distribution, generate the deliverable radiotherapy plan (31) and send the deliverable radiotherapy plan (31) to a predefined radiotherapy delivery system for final verification and execution.