Markerless Anatomical Tracking Using Patient-Specific AI Guidance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current real-time image-guided radiation therapy systems require the implantation of fiducial markers, which are costly, time-consuming, risky, and prone to tracking errors, limiting their accessibility and effectiveness, especially for patients who cannot tolerate marker implantation.
Innovation Solution
A markerless approach using a patient-specific, individually trained conditional Generative Adversarial Network (cGAN) for real-time image guidance, which leverages pre-treatment physician-drawn contours and CT scans to accurately track anatomical objects without fiducial markers, enabling real-time motion monitoring and treatment adjustment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If fiducial markers are implanted for real-time tracking, then measurement precision is improved, but device complexity and risk increase
Solution Approach 1:
The invention extracts the tracking function from physical fiducial markers and implements it through software-based artificial intelligence image analysis. The system removes the need for marker implantation by using automated detection of anatomical landmarks in kV images, thereby eliminating the complexity and risk associated with surgical marker placement while maintaining tracking precision.
Solution Approach 2:
The invention replaces the mechanical/surgical system of fiducial marker implantation with an automated computational system using artificial intelligence. Instead of physically inserting markers into the patient's body, the system uses machine learning algorithms to automatically identify and track anatomical structures in real-time images, substituting a mechanical procedure with a software-based solution.
2Measurement precision
If fiducial markers are implanted for real-time tracking, then measurement precision is improved, but loss of time increases
Solution Approach 1:
The system performs preliminary training of the artificial intelligence model using pre-treatment CT scans and physician-drawn contours before the actual treatment begins. This preliminary action allows the AI to learn patient-specific anatomical features in advance, so that during real-time treatment, the system can immediately begin tracking without requiring any additional setup time or marker implantation procedures.
3Measurement precision
If fiducial markers are implanted for real-time tracking, then measurement precision is improved, but loss of substance increases
Solution Approach 1:
The invention replaces expensive, permanent fiducial markers with a cost-effective software-based solution that uses standard kV imaging equipment already available in treatment rooms. The AI system processes regular treatment images to extract tracking information, eliminating the need for costly marker materials and their associated implantation procedures, thereby significantly reducing overall treatment costs.
4Measurement precision
If fiducial markers are implanted for real-time tracking, then measurement precision is improved, but reliability worsens
Solution Approach 1:
The system implements continuous feedback by repeatedly analyzing successive kV images during treatment to track anatomical motion in real-time. The AI model continuously updates the position of anatomical landmarks based on each new image, providing ongoing feedback about target motion that allows for dynamic adaptation of treatment delivery, thereby improving reliability through continuous verification and adjustment.
Data Source
AI summary
An image guidance method for treatment by a medical device. The method comprises imaging a target area to which the treatment is to be delivered. During the interventional procedure, an image from the imaging is analysed by a patient-specific, individually trained artificial neural network to determine the position of at least one or more anatomical objects of interest present in the target area. The determined position(s) is output to the medical device for the delivery of treatment.