Method and apparatus for enhanced X-ray computing arrays

Inventors

Kahn, Paul AnthonyEllenor, Christopher WilliamFunk, TobiasKonings, Oleg John

Assignees

Triple Ring Technologies Inc

Publication Number

US-10231687-B2

Publication Date

2019-03-19

Expiration Date

2035-10-16

Interested in licensing this patent?

MTEC can help explore whether this patent might be available for licensing for your application.


Abstract

An x-ray imaging system utilizes enhanced computing arrays. A plurality of x-ray illumination source positions are utilized to produce x-ray radiation at each of the x-ray illumination source positions and to project x-ray radiation towards an object. A detector detects x-ray radiation from the object and transmits detector images for each of the illumination source positions. A memory buffer stores the detector images from the detector. A graphics processing unit formats the detector images and constructs a complete frame data set with the detector images for each of the illumination source positions. Another graphics processing unit receives the complete frame data set and performs image reconstruction on the complete frame data set.

Core Innovation

The invention provides a method and apparatus for enhanced x-ray imaging using advanced computing arrays. Multiple x-ray illumination source positions generate x-ray radiation directed towards an object. A detector captures the x-ray radiation transmitted or scattered from the object, producing detector images for each of the source positions. These detector images are stored in a memory buffer, where a first graphics processing unit (GPU) formats and constructs a complete frame data set comprising the images from all source positions.

The complete frame data set is then transmitted to a second GPU, which performs image reconstruction operations, using the accumulated data to create a final image. The system architecture enables efficient real-time image processing by offloading pre-processing tasks to a first GPU and dedicating a second GPU to computationally intensive reconstruction tasks, thereby maximizing GPU resource usage and minimizing performance losses. The data transfer between GPUs is executed in a single burst operation, and the system supports features like direct memory access and high-speed data transfer rates.

The problem addressed by the invention is the need for a cost-effective imaging and display system capable of processing high quantities of image data at high frame rates in real-time environments, especially in cases such as imaging moving objects like beating hearts or during surgical procedures. Traditional GPU-based systems suffer from performance losses when operating continuously, sometimes requiring additional GPUs and increasing system complexity and cost. By partitioning processing between a pre-processing GPU and reconstruction GPU(s), the system achieves high quality imaging, supports multiple illumination focal spot positions, and meets strict speed and performance requirements for real-time medical imaging.

Claims Coverage

The patent contains two independent claims that define the main inventive features of the x-ray imaging system and method.

X-ray imaging system utilizing multiple GPUs for real-time image reconstruction

The system comprises: - A plurality of x-ray illumination source positions for generating and projecting x-ray radiation towards an object. - A detector that captures x-ray radiation and generates detector images for each illumination position. - A memory buffer for storing these detector images. - A first graphics processing unit coupled to the memory buffer, responsible for formatting detector images and constructing a complete frame data set. - A second graphics processing unit, coupled to the first, that receives the complete frame data set and executes image reconstruction on it. - A bus connecting both GPUs, transmitting the complete frame data set in a single burst operation.

Method for x-ray imaging using GPU partitioning for frame construction and reconstruction

The method entails: 1. Generating x-ray radiation from multiple illumination source positions and directing it to the object. 2. Detecting incoming x-rays and creating detector images for each source position. 3. Employing a first graphics processing unit to construct a complete frame data set with the detector images. 4. Transmitting the complete frame data set from the first GPU to a second GPU in a single burst operation. 5. Using the second GPU to conduct image reconstruction on the received complete frame data set.

In summary, the inventive features include a specialized architecture dividing pre-processing and image reconstruction tasks between dedicated GPUs, enabling fast and efficient transfer and processing of x-ray detector image data for high frame-rate, real-time imaging.

Stated Advantages

Reduces performance losses and maximizes GPU utilization by separating pre-processing and reconstruction tasks between multiple GPUs.

Enables cost-effective real-time processing and production of high quality images at high frame rates for medical imaging applications.

Supports efficient handling of large quantities of imaging data from multiple x-ray illumination source positions, meeting the demands of dynamic and high-speed imaging scenarios.

Reduces system cost and complexity by preventing the need for additional GPUs to overcome traditional performance bottlenecks.

Documented Applications

Medical x-ray imaging, including tomosynthesis and computed tomography requiring multiple image angles.

Imaging of moving objects, such as beating hearts, in applications like cardiac catheterization, angioplasty, and ablation.

Surgical environments needing real-time imaging at high frame rates.

Low-dose fluoroscopic systems with large field of view for interventional radiology.

Scanning-Beam Digital X-ray (SBDX) fluoroscopy for cardiac imaging.

JOIN OUR MAILING LIST

Stay Connected with MTEC

Keep up with active and upcoming solicitations, MTEC news and other valuable information.