20240041412. FEW-VIEW CT IMAGE RECONSTRUCTION SYSTEM simplified abstract (RENSSELAER POLYTECHNIC INSTITUTE)
FEW-VIEW CT IMAGE RECONSTRUCTION SYSTEM
Organization Name
RENSSELAER POLYTECHNIC INSTITUTE
Inventor(s)
Ge Wang of Loudonville NY (US)
Wenxiang Cong of Albany NY (US)
FEW-VIEW CT IMAGE RECONSTRUCTION SYSTEM - A simplified explanation of the abstract
This abstract first appeared for US patent application 20240041412 titled 'FEW-VIEW CT IMAGE RECONSTRUCTION SYSTEM
Simplified Explanation
The abstract describes a system for few-view computed tomography (CT) image reconstruction. The system consists of a preprocessing module, a first generator network, and a discriminator network. The preprocessing module applies a ramp filter to an input sinogram to obtain a filtered sinogram. The first generator network learns a filtered back-projection operation using the filtered sinogram and generates a first reconstructed image as output, which corresponds to the input sinogram. The discriminator network determines whether a received image corresponds to the first reconstructed image or a ground truth image. The generator and discriminator networks are part of a Wasserstein Generative Adversarial Network (WGAN) and are optimized using an objective function based on Wasserstein distance and gradient penalty.
- The system reconstructs CT images using a few-view approach.
- A preprocessing module applies a ramp filter to the input sinogram to obtain a filtered sinogram.
- The first generator network learns a filtered back-projection operation and generates a reconstructed image.
- The discriminator network determines if a received image corresponds to the reconstructed image or a ground truth image.
- The generator and discriminator networks are part of a WGAN and are optimized using an objective function based on Wasserstein distance and gradient penalty.
Potential applications of this technology:
- Medical imaging: The system can be used in medical CT scanners to reconstruct high-quality images with fewer views, reducing radiation exposure for patients.
- Industrial inspection: The system can be applied in non-destructive testing to reconstruct images of objects with limited views, improving inspection accuracy.
Problems solved by this technology:
- Limited-view CT reconstruction: The system addresses the challenge of reconstructing high-quality CT images using a limited number of views, which is useful in scenarios where complete views are not feasible.
- Image quality and accuracy: By incorporating a WGAN and optimizing with Wasserstein distance and gradient penalty, the system improves the quality and accuracy of reconstructed images compared to traditional methods.
Benefits of this technology:
- Reduced radiation exposure: Few-view CT reconstruction reduces the number of views required, resulting in lower radiation exposure for patients during imaging procedures.
- Improved efficiency: The system enables faster image reconstruction by utilizing a few-view approach, increasing the efficiency of CT scanning.
- Enhanced image quality: By incorporating a WGAN and optimizing with Wasserstein distance and gradient penalty, the system produces high-quality CT images with improved accuracy and reduced artifacts.
Original Abstract Submitted
a system for few-view computed tomography (ct) image reconstruction is described. the system includes a preprocessing module, a first generator network, and a discriminator network. the preprocessing module is configured to apply a ramp filter to an input sinogram to yield a filtered sinogram. the first generator network is configured to receive the filtered sinogram, to learn a filtered back-projection operation and to provide a first reconstructed image as output. the first reconstructed image corresponds to the input sinogram. the discriminator network is configured to determine whether a received image corresponds to the first reconstructed image or a corresponding ground truth image. the generator network and the discriminator network correspond to a wasserstein generative adversarial network (wgan). the wgan is optimized using an objective function based, at least in part, on a wasserstein distance and based, at least in part, on a gradient penalty.
