Automatic Patient Positioning Using Multi-Sensor Pose Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current minimally invasive robotic surgical systems face challenges in accurately positioning medical equipment relative to a patient's body, particularly in procedures requiring precise targeting of radiation, due to limitations in sensor data processing and lack of automatic adjustment mechanisms.
Innovation Solution
A system utilizing multiple sensors, including RGB, depth, and infrared sensors, processes sensor data through a machine learning model to determine patient pose and shape characteristics, calculates adjustment parameters, and adjusts medical equipment, such as an articulated robotic arm or patient bed, for precise positioning.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual positioning methods are used for medical equipment, then device complexity is reduced, but positioning precision and targeting accuracy deteriorate
Solution Approach 1:
The patent replaces manual mechanical positioning with an automated computer vision system that uses sensors, machine learning models, and robotic actuators to automatically position medical equipment. The system captures images, processes them through trained models to determine patient pose and shape characteristics, and automatically adjusts equipment position based on calculated adjustment parameters, eliminating the need for manual positioning while achieving high precision.
Solution Approach 2:
The system performs self-positioning by automatically capturing images, processing them through the machine learning model, determining patient characteristics, calculating adjustment parameters, and executing position adjustments without requiring manual intervention. The robotic system serves itself by integrating perception, decision-making, and execution capabilities within a single automated workflow.
2Productivity
If automatic positioning systems are implemented, then positioning speed and accuracy are improved, but device complexity increases
Solution Approach 1:
The patent divides the automatic positioning system into distinct functional modules: image capture sensors, a machine learning model for processing images and determining patient characteristics, a calculation module for computing adjustment parameters, and execution mechanisms for position adjustment. This segmentation allows each component to be optimized independently while working together to achieve high positioning efficiency.
Solution Approach 2:
The system performs preliminary actions by capturing images and processing them through a pre-trained machine learning model before the actual positioning operation. The model is trained in advance on patient data to recognize pose and shape characteristics, enabling rapid and accurate positioning when the system is deployed, thus improving overall efficiency.
3Object-affected harmful factors
If precise targeting is achieved through automatic positioning, then irradiation of healthy tissue is minimized, but measurement and detection difficulty increases
Solution Approach 1:
The patent replaces difficult manual detection and measurement of patient pose and shape with an automated computer vision system. Multiple sensors capture images from different views, and a machine learning model automatically processes these images to determine patient characteristics, eliminating the need for manual measurement while achieving high accuracy in detecting pose and shape.
Solution Approach 2:
The machine learning model serves as an intermediary between the raw sensor data and the final positioning decision. The model translates complex multi-view images into simplified patient pose and shape characteristics, making the detection process more manageable and accurate while reducing the difficulty of direct measurement.
4Measurement precision
If multi-view data fusion is used for patient characterization, then positioning accuracy is improved, but data processing complexity increases
Solution Approach 1:
The patent segments the data processing task by having different sensors capture images from different views, which are then processed independently through the machine learning model to extract specific patient characteristics. This segmentation of the imaging and processing tasks makes the complex data fusion more manageable while improving accuracy through multiple perspectives.
Data Source
AI summary
Disclosed is a method and a system for automatic positioning of a medical equipment with respect to a patient. The method includes obtaining sensor data related to the patient, from a plurality of sensors fixed relative to the medical equipment. The method further includes processing the sensor data to determine at least one pose characteristic of the patient and at least one shape characteristic of the patient. The method further includes determining at least one adjustment parameter for the medical equipment based on the at least one pose characteristic of the patient and the at least one shape characteristic of the patient. The method further includes adjusting the medical equipment based on the at least one adjustment parameter.