Back to results
US 2024/0389955 A1Published application
Hybrid method for fast functional imaging with sparse sampling in tomography
OwnerPeng Hu, Lihong Wanginventors
Life of this patent
Estimating the term…
Abstract
Certain aspects pertain to methods and systems for compressing tomographic imaging system matrices and/or functional tomographic imaging.
Claims29
A tomographic imaging system matrix compression method, comprising:
- (a) acquiring a plurality of detected tomographic system responses;
- (b) shifting one or more of the detected tomographic system responses to align arrival times;
- (c) for each virtual sensor of a plurality of virtual sensors, determining an approximate system response that is a linear combination of a plurality of basis functions, wherein the approximate system response is determined by performing a singular value decomposition (SVD) process on the aligned, detected tomographic system responses to obtain a plurality of coefficients for the linear combination;
- and
- (d) determining a compressed system matrix including approximate system responses for all virtual sensors from the plurality of coefficients determined for each of the virtual sensors.
2 more elements
(canceled)
7-8. (canceled)
(canceled)
15-16. (canceled)
A system comprising:
- a tomographic imaging device comprising one or more sensor arrays;
- one or more processors in electrical communication with the tomographic imaging device;
- and
- one or more processor-readable media storing instructions which, when executed by the one or more processors, cause performance of:
- (a) performing piece-wise compression of a plurality of aligned system responses to generate a compressed system matrix;
- (b) applying the compressed system matrix to generate a plurality of simulated virtual sensor signals;
- and
- (c) performing an iterative reconstruction procedure to reconstruct tomographic image from detected sensor signals by comparing a difference between the simulated virtual sensor signals and the detected sensor signals.
5 more elements
(canceled)
(canceled)
A method for functional tomographic imaging, the method comprising:
- (a) performing iterative reconstruction of a smooth modulation image using a set of sparsely sampled functional signals from a plurality of sensors of a tomographic imaging device;
- (b) obtaining a modulated prior image by upsampling the smooth modulation image and multiplying the upsampled smooth modulation image by a densely sampled prior image;
- (c) applying a dense-sampling system matrix to the modulated prior image to determine a dense forward solution and reconstructing a modulated universal back-projection (UBP) image from the dense forward solution using a UBP method;
- (d) applying a sparse sampling system matrix to the modulated prior image to determine a sparsely sampled solution and reconstructing a residual image using the UBP method from a difference between the set of sparsely sampled functional signals and the sparsely sampled solution;
- and
- (e) determining a hybrid functional image by adding the modulated UBP image to the residual image.
3 more elements
28-32. (canceled)
A system comprising:
- a tomographic imaging device comprising one or more sensor arrays;
- one or more processors in electrical communication with the tomographic imaging device;
- and
- one or more processor-readable media storing instructions which, when executed by the one or more processors, cause performance of:
- (a) performing iterative reconstruction of a smooth modulation image using a set of sparsely sampled functional signals from a plurality of sensors of a tomographic imaging device;
- (b) obtaining a modulated prior image by upsampling the smooth modulation image and multiplying the upsampled smooth modulation image by a densely sampled prior image;
- (c) applying a dense-sampling system matrix to the modulated prior image to determine a dense forward solution and reconstructing a modulated universal back-projection (UBP) image from the dense forward solution using a UBP method;
- (d) applying a sparse sampling system matrix to the modulated prior image to determine a sparsely sampled solution and reconstructing a residual image using the UBP method from a difference between the set of sparsely sampled functional signals and the sparsely sampled solution;
- and
- (e) determining a hybrid functional image by adding the modulated UBP image to the residual image.
7 more elements
(canceled)
Dependencies are inferred from the claim text.