Multi-Resolution Detector with Deep Learning for CT Bandwidth
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Computed tomography scanners face a trade-off between achieving high spatial resolution and increasing hardware costs and communication bandwidth demands, particularly with multi-slice CT scanners, where a finer detector resolution requires more expensive hardware and higher data transfer rates.
Innovation Solution
Implementing a multi-resolution detector with finer pixel pitch near the center of the field of view and coarser pixel pitch at the periphery, combined with a deep learning neural network to enhance resolution post-data transmission, thereby maintaining high resolution without increasing communication bandwidth requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If detector pixel size is decreased to improve spatial resolution, then resolution is improved, but detector cost and communication bandwidth requirements increase
Solution Approach 1:
The detector employs variable pixel pitch across different regions: fine pixel pitch (e.g., 0.5mm) in the central region where high spatial resolution is critical for visualizing fine details, and coarse pixel pitch (e.g., 1.0mm) in the peripheral regions where resolution requirements are lower. This local differentiation maintains measurement precision where needed while reducing overall detector complexity and cost.
Solution Approach 2:
The detector array is segmented into multiple regions with different resolution characteristics. The central region uses finer pixel pitch for high-resolution imaging, while peripheral regions use coarser pixel pitch. This segmentation allows the system to allocate computational and hardware resources efficiently, improving overall spatial resolution performance without uniformly increasing detector complexity across the entire field of view.
2Measurement precision
If more detector pixels are used to cover the same area, then resolution is improved, but data transmission requirements increase
Solution Approach 1:
The system transmits only the necessary amount of data from each detector region based on its resolution requirements. Central regions with fine pixel pitch transmit higher-resolution data, while peripheral regions with coarse pixel pitch transmit lower-resolution data. This local quality differentiation reduces the total data volume requiring transmission while maintaining spatial resolution where critical.
Solution Approach 2:
The system applies fine pixel pitch (excessive resolution) only to the central region where it is necessary for accurate imaging, while using coarse pixel pitch (insufficient resolution) in peripheral regions where the same level of detail is not required. This partial application of high resolution optimizes the balance between image quality and data transmission requirements.
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
This approach allows for improved spatial resolution in critical areas while reducing the overall number of detector elements and communication bandwidth, thus lowering the cost of the CT scanner without sacrificing image quality.
Implementation Method 1
A radiation source, such as an X-ray tube, irradiates the body from one side
Implementation Method 2
At least one detector on the opposite side of the body receives radiation transmitted through the body substantially in the projection volume. The attenuation of the radiation that has passed through the body is measured by processing electrical signals received from the detector.
Data Source
AI summary
A method and apparatus is provided that uses a deep learning (DL) network together with a multi-resolution detector to perform X-ray projection imaging to provide improved resolution similar to a single-resolution detector but at lower cost and less demand on the communication bandwidth between the rotating and stationary parts of an X-ray gantry. The DL network is trained using a training dataset that includes input data and target data. The input data includes projection data acquired using a multi-resolution detector, and the target data includes projection data acquired using a single-resolution, high-resolution detector. Thus, the DL network is trained to improve the resolution of projection data acquired using a multi-resolution detector. Further, the DL network is can be trained to additional correct other aspects of the projection data (e.g., noise and artifacts).