Linxi Shi

Title: Linxi Shi and a Dual-Domain Neural Network for 3D Medical Imaging

Introduction

Linxi Shi is an inventor based in Oakland, California, associated with Neuraltrak, Inc. His patent work addresses medical 3D imaging reconstruction from limited-angle 2D image data, with the goal of reducing the amount of data required compared with traditional scanner-based reconstruction.

Latest Patents

Shi is associated with 1 patent: “System and Method for 3D Imaging Reconstruction Using Dual-Domain Neural Network.” The application describes capturing 2D projections of a target across a limited angle, calibrating those projections, and generating a sinogram from the calibrated data. A dual-domain neural network then processes the sinogram to produce a geometrically corrected and calibrated 3D volumetric image.

According to the provided description, the approach requires only 30–50% of the data used by traditional scanner-based 3D reconstruction. Reducing the required acquisition data may lower radiation exposure and improve clinical efficiency.

Career Highlights

Shi’s listed invention applies neural-network methods to medical imaging reconstruction. Its workflow combines projection calibration, sinogram generation, and dual-domain processing to create a 3D image from limited-angle measurements.

Collaborations

Shi is associated with Neuraltrak, Inc. His listed coworker is Shiyu Xu.

Conclusion

Linxi Shi’s patent work focuses on using a dual-domain neural network to reconstruct calibrated 3D medical images from limited-angle 2D projections, with the potential to reduce data requirements and radiation exposure.