Non-Imaging Transcranial Ultrasound Probe Positioning With MRI-Guided AI
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current transcranial ultrasound systems face challenges in accurately targeting anatomical structures in the brain due to varying acoustic properties of skull bone, air cavities, and other tissues, leading to sub-optimal ultrasound delivery and potential side effects, which complicates the operating process and increases system complexity.
Innovation Solution
Employing domain-specific large vision models (DSLVM) to assist in positioning and segmenting ultrasound probes, using pre-procedure MRIs to enhance the placement of non-imaging probes, and providing real-time guidance through neuro-navigation and beamforming adjustments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If operator manually adjusts probe placement and positioning, then system complexity is reduced, but targeting accuracy deteriorates due to varying acoustic properties of skull bone, air cavities, and tissues
Solution Approach 1:
An AI model acts as an intermediary between the operator and the ultrasound system, automatically analyzing MRI scans to identify anatomical structures (skull bone, air cavities, tissues) and calculate optimal probe placement. This intermediary process resolves the contradiction by providing expert-level targeting accuracy without requiring the operator to manually compensate for complex acoustic variations, thus maintaining high precision while keeping the user interface relatively simple.
Solution Approach 2:
The system performs preliminary analysis of the patient's anatomy using MRI scans before the actual ultrasound procedure. The AI model pre-identifies all relevant anatomical structures and pre-calculates the optimal probe positioning and beamforming parameters. This preliminary action ensures accurate targeting is achieved before the operator even places the probe, resolving the contradiction by embedding the complex analysis work in advance rather than requiring real-time manual adjustment during the procedure.
2Measurement precision
If AI model automatically identifies anatomical structures and optimizes probe placement, then targeting accuracy improves, but device complexity increases
Solution Approach 1:
The AI model performs self-service by automatically analyzing MRI scans, identifying anatomical structures, and determining optimal probe placement without requiring manual intervention for each parameter. The system serves itself by generating beamforming parameters and positioning recommendations autonomously. This self-service capability improves probe placement accuracy while managing complexity by automating the analytical tasks rather than requiring complex manual adjustment mechanisms.
Solution Approach 2:
The patent replaces manual mechanical adjustment of probe placement with an AI-based computational system that analyzes MRI data and automatically determines optimal positioning. This substitution of mechanical/manual processes with intelligent algorithms improves placement accuracy by considering all acoustic properties simultaneously, while the complexity is managed through software-based solutions rather than complex mechanical adjustment mechanisms.
3Reliability
If operator must avoid air cavities and adjust for each structure type, then ultrasound delivery reliability improves, but ease of operation deteriorates
Solution Approach 1:
The AI model provides feedback to the operator by automatically identifying air cavities and other anatomical structures that could compromise ultrasound delivery, and by recommending alternative probe placements that avoid these obstacles. This feedback mechanism ensures reliable ultrasound delivery by preventing propagation through air cavities while maintaining ease of operation through automated guidance rather than requiring the operator to manually identify and avoid each problematic structure.
Solution Approach 2:
The system performs preliminary anti-action by proactively identifying air cavities and sub-optimal ultrasound paths before the procedure begins, and by pre-calculating alternative probe placements that avoid these obstacles. This preliminary protective action ensures ultrasound delivery reliability by preventing propagation through harmful structures while simplifying operation through advance planning rather than requiring real-time manual avoidance maneuvers.
4Reliability
If system provides real-time guidance and beamforming adjustments, then treatment reliability improves, but device complexity increases
Solution Approach 1:
The system performs preliminary calculation of beamforming parameters and probe positioning based on MRI analysis before the actual ultrasound treatment begins. By pre-determining the optimal beamforming settings and probe placement, the system ensures treatment reliability without requiring complex real-time adjustments during the procedure. The complexity is managed by performing the analytical work in advance, allowing the treatment phase to proceed with pre-established reliable parameters.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Improves the accuracy and efficacy of ultrasound delivery to targeted brain anatomy by reducing operator complexity and ensuring optimal probe placement, thereby enhancing treatment reliability and safety.
Implementation Method 1
ultrasonic probe, a control subsystem in communication with an input device
Implementation Method 2
These structures have different acoustic properties, which refract, reflect, and diffract the ultrasound field
Implementation Method 3
These structures have different acoustic properties, which refract, reflect, and diffract the ultrasound field
Implementation Method 4
These structures have different acoustic properties, which refract, reflect, and diffract the ultrasound field
Data Source
AI summary
Transcranial ultrasound systems (TUS) and methods use domain-specific large vision models (DSLVM) artificial intelligence systems to improve the efficacy of non-imaging probes. A positioning DSLVM assists an operator in improving the placement of the probe on a patient's head. The positioning DSLVM uses a pre-procedure MRI of the patient's head, target dose plan, target anatomy, and the probe's position information on the scalp, and it outputs the control parameters for the probe's beamformer. A segmenting DSLVM helps an operator with the optimal initial placement of the non-imaging probe by highlighting anatomical structures in color.