Reconstruction From Projection TRIZ Cases

Direct derivative calculation on filtering lines eliminates interpolation errors in x-ray image reconstruction.

A tomosynthesis system generates synthetic projection images to align low and high energy scan angles for accurate difference image reconstruction.

Dynamic 3D reconstruction identifies regions of interest to guide targeted re-acquisitions, reducing radiation dosage while maintaining diagnostic precision.

Patch-wise computations in a learned iterative scheme reduce memory requirements while correcting streaking artifacts in cone beam computed tomography.

A multi-modality imaging system combines X-ray CT, neutron CT, and electromagnetic tomography with machine learning to estimate material properties in large rock cores.

Calculates material-specific noise from attenuation data to correct beam hardening artifacts during iterative tomography.

Reprojecting 3D markers in standard projection images confirms radiological signs while reducing access time to clinical information.

A framelet-based iterative maximum-likelihood algorithm reconstructs spectral images using standard energy-integrating detectors.

Residual convolutional neural networks estimate physiological artifacts in multi-shot echo planar imaging sequences to reconstruct high isotropic resolution images.

Viewing point conversion isolates relative internal body motion from proximal and distant tissue interference for precise heart function evaluation.

Replacing k-edge artifacts in spectral attenuation curves with approximations improves signal-to-noise ratio and effective atomic number estimates.

Information processing apparatus performs basis conversion on signal data to calculate multiple bases.

Merges circular trajectory efficiency with linear trajectory completeness to eliminate cone beam artifacts in medical imaging.

Gradient stiffness catheters navigate tortuous vessels while maintaining stability, reducing tissue damage and procedural time.

A poly-energetic iterative Filtered Backprojection algorithm decomposes attenuation coefficients to reconstruct artifact-free cross-sectional images.

Deep learning algorithms reconstruct tomographic images from raw data, eliminating noise and artifacts without iterative optimization.

Segmenting event data into portions allows the system to generate intermediate images quickly while refining composite data for better quality.

Combining a neural network with compressed sensing reconstructs high-quality MRI images from undersampled data, mitigating artifacts and noise.

A method processes four-dimensional angiography data sets by establishing static time parameter sets and applying window functions to create mask data sets.

Rendering algorithms enhance tissue visibility in 2D pre-scan projection images for computed tomography systems.

A medical image processing apparatus generates color image data from time series contrast agent signals to visualize blood vessel dynamics.

A magnetic resonance imaging method removes k space values exceeding threshold criteria to generate cleaner measurement datasets.

Reconstructs parametric images from continuous bed motion PET data using recorded position-time coordinates.

Machine learning discriminates border regions to generate a confidence map that indicates missing projection directions and reduces limited angle artifacts.