System and method for reconstruction of 3D volumes from biplanar radiographic images
AssigneeSee All AI Inc.
InventorEugene Alma Gregerson
Life of this patent
Estimating the term…
Abstract
A system and method combine optical and radiographic data to enhance imaging capabilities. Specifically, the system combines visually obtained patient pose position information and radiographic image information to facilitate calibrated surgical navigation. A reconstruction of a 3D CT volume is generated from biplanar X-ray projections which are back projected into two separate volumes and then concatenated into a single volume along a new dimension and passed through a pretrained deep learning model to decode the concatenated volume into a single 3D volume.
Claims5
A method for constructing a three-dimensional (3D) volume from a set of two or more medical images, the method comprising:
- A) acquiring two or more biplanar radiographic images with an X-ray imaging device;
- B) calibrating each of the biplanar radiographic images both intrinsically and extrinsically in relation to a common coordinate system;
- C) encoding the calibrated images using a machine learning or deep learning algorithm executing on a processor;
- D) back-projecting the encoded images into separate respective volumes and concatenating the separate respective volumes into a combined volume along a new dimension within the common coordinate system based on known relative 3D poses;
- and
- E) decoding the back-projected combined volume to reconstruct the 3D volume using a pretrained 3D neural network.
A computer program product comprising a non-transitory computer-readable medium storing instruction that, when executed by a processor, cause the processor to perform a method for constructing a three-dimensional (3D) volume from a set of two or more medical images, the method comprising:
- A) acquiring two or more biplanar radiographic images with an X-ray imaging device;
- B) calibrating each of the biplanar radiographic images both intrinsically and extrinsically in relation to a common coordinate system;
- C) encoding the calibrated images using a machine learning or deep learning algorithm executing on a processor;
- D) back-projecting the encoded images into separate respective volumes and concatenating the separate respective volumes into a combined volume along a new dimension within the common coordinate system based on known relative 3D poses;
- and
- E) decoding the back-projected combined volume to reconstruct the 3D volume using a pretrained 3D neural network.
Dependencies are inferred from the claim text.