CT Tube Current Calculation Using ML and Scout Images
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current CT scanning methods face errors in determining the tube current value due to the differences between the actual human body and equivalent phantoms, leading to inaccuracies in noise levels of medical images.
Innovation Solution
A method using a trained machine learning model to calculate the tube current value based on scout images, scanning protocols, and preset image noise parameters, allowing for an updated scanning protocol to be applied during the CT scan.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If an equivalent phantom is used to determine tube current, then the exposure control can be standardized, but errors occur due to differences between actual human body and phantom structure
Solution Approach 1:
The patent uses a scout image (a low-dose preliminary CT scan) to create a digital copy of the patient's actual anatomy, which then serves as the basis for tube current calculation. This digital copy captures the real patient's body structure, tissue density distribution, and anatomical variations, replacing the need for physical phantoms while maintaining measurement accuracy.
Solution Approach 2:
The system calculates tube current by analyzing multiple parameters extracted from the scout image, including body thickness, tissue density distribution, and attenuation coefficients. These parameters are dynamically adjusted based on the actual patient anatomy rather than using fixed phantom-based values, enabling precise adaptation to each patient's unique structure.
2Reliability
If laboratory phantom measurement is used to establish correspondence, then a reference system can be created, but errors remain due to the gap between phantom and actual human body
Solution Approach 1:
The system performs self-calibration by using the patient's own scout image to automatically determine the appropriate tube current. The algorithm extracts anatomical features directly from the patient's imaging data and adjusts exposure parameters accordingly, eliminating the need for external phantom-based calibration and its associated errors.
Solution Approach 2:
The patent replaces the mechanical/physical phantom-based measurement system with a computational imaging system. Instead of using physical phantoms and manual measurements, the system uses digital image processing, automated feature extraction, and algorithm-based tube current calculation to achieve more accurate and consistent results.
3Manufacturing precision
If different scan parameters are used for different scanning sites, then image quality can be optimized, but the complexity of parameter selection increases
Solution Approach 1:
The system performs a preliminary scout scan before the actual CT examination to gather anatomical information about the patient's body structure at the specific scanning site. Based on this preliminary data, the algorithm automatically determines the optimal scan parameters including tube current, voltage, and rotation speed, eliminating the need for manual parameter selection by the operator.
Solution Approach 2:
The system uses feedback from the scout image analysis to automatically adjust scan parameters. The algorithm continuously monitors anatomical features extracted from the scout image and adjusts the tube current and other parameters accordingly, ensuring optimal image quality for each specific scanning site without requiring complex manual intervention.
Data Source
AI summary
The present application provides a method for obtaining a tube current value, a medical imaging system, and a non-transitory computer-readable storage medium. The example method for obtaining a tube current value includes obtaining a scanning protocol, performing a scout scan to obtain a scout image of a subject under examination, and obtaining a tube current value on the basis of a trained machine learning model, according to the scout image, the scanning protocol, and a preset image noise parameter.