Neural Network Procedure Parameters for Endovascular Implant Positioning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Endovascular occlusion of aneurysms and vessel malformations is complex, time-consuming, and risky, often requiring multiple attempts to correctly position microcatheters and implants due to geometric and hemodynamic challenges, with risks including thromboembolic events and overpacking.
Innovation Solution
A method using trained neural networks to analyze intraprocedural projection maps and simulate the positioning of medical objects within hollow organs, providing procedure parameters for improved selection and control of microcatheters and implants, assisted by robotic systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional manual positioning methods are used for microcatheter and implant placement, then the interventional procedure can be performed with existing equipment, but the procedure becomes complex and time-consuming requiring multiple attempts
Solution Approach 1:
The patent creates a virtual copy (digital twin) of the patient's anatomy from pre-procedural imaging data. This virtual model allows simulation and planning of the entire interventional procedure before actual execution, enabling optimization of microcatheter and implant positioning without repeated physical attempts during the procedure.
Solution Approach 2:
The system performs preliminary simulation and planning using the virtual anatomical model before the actual interventional procedure. Procedure parameters such as microcatheter trajectory, implant positioning, and deployment timing are predetermined through virtual experimentation, eliminating the need for multiple corrective attempts during the live procedure.
2Manufacturing precision
If multiple attempts are made to position medical objects correctly, then positioning accuracy can be improved, but the procedure time and complexity increase
Solution Approach 1:
A virtual copy of the patient's anatomy is created from pre-procedural imaging, allowing repeated simulation of positioning scenarios without time cost. The virtual environment enables exhaustive testing of different microcatheter trajectories and implant positions to identify the optimal solution before actual procedure execution.
Solution Approach 2:
Positioning accuracy is determined in advance through virtual simulation and planning. The system calculates optimal microcatheter insertion paths and implant deployment parameters beforehand, so that during the actual procedure, the operator can execute the pre-planned trajectory directly without time-consuming trial and error.
3Reliability
If the interventional procedure is performed manually without assistance, then equipment requirements are minimized, but the success rate decreases due to geometric and hemodynamic challenges
Solution Approach 1:
The system introduces an intermediary computational layer between pre-procedural imaging and intra-procedural action. This intermediary virtual model processes anatomical data, simulates hemodynamic effects, and predicts optimal positioning parameters, serving as a decision-support mediator that increases success rate without requiring complex robotic or automated equipment during the actual procedure.
Solution Approach 2:
The patent replaces complex real-time mechanical adjustment during the procedure with pre-computed virtual simulation. Instead of manually trial-and-error positioning assisted by simple imaging, the system uses virtual physics-based simulation of hemodynamic effects and tissue interaction to determine optimal positioning, substituting mechanical complexity with computational modeling.
4Productivity
If real-time guidance systems are implemented, then positioning accuracy and efficiency improve, but the procedural complexity and equipment requirements increase
Solution Approach 1:
All guidance and positioning information is prepared in advance through virtual simulation using pre-procedural imaging data. The system performs real-time guidance by retrieving pre-computed optimal trajectories and parameters from the virtual model, avoiding the need for complex real-time sensing, actuation, and control systems during the actual procedure.
Data Source
AI summary
A method for providing at least one procedure parameter includes acquiring at least one intraprocedural projection map that maps a hollow organ of an examination object with at least one medical object positioned in the hollow organ. A trained function is applied to the at least one intraprocedural projection map as input data. At least one parameter of the trained function is adjusted based on a simulation of a virtual positioning of at least one medical training object in a training hollow organ and of hemodynamics in the training hollow organ that are influenced by the at least one medical training object. The at least one procedure parameter is provided as output data of the trained function. The at least one procedure parameter includes a movement specification for the at least one medical object.