Reconstruction From Projection TRIZ Cases

A photon counting x-ray detector captures individual photons during continuous source rotation to generate high-resolution tomographic projections.

Voxelizing spatial data and classifying features bridges GPS unavailability in urban environments to achieve centimeter-level localization accuracy.

A recurrent neural network determines missing k-space data from limited samples to reconstruct high-quality MRI images.

A hybrid cone beam image reconstruction method using ray-wise 3D weighting dependent on helical pitch and z-distance.

Multi-grid iterative reconstruction accelerates tomographic image processing by applying coarse-to-fine filtering, reducing computational time for nano CT.

Cross-modal deep learning engine generates super-resolution tomographic images from holographic data.

Helical projection data rebinning aligns orthogonal source-detector pairs in dual-source computed tomography systems.

Blends analytic and iterative image reconstruction components in frequency space to reduce noise while preserving texture.

A PET reconstruction system groups detector units by functional status to generate accurate images from valid data.

A medical image processing apparatus automatically determines composition ratios for dual energy images to generate composite diagnostic views.

Calculating local non-uniformity in modified cumulative distribution functions detects spatial artifacts invisible to global histogram analysis.

A voxel-dependent update factor weights backprojected errors during iterative tomographic reconstruction to improve image quality.

Reconstructs virtual machine images from equivalent blocks on local or donor devices, reducing network traffic and transfer time during provisioning.

A time measuring unit tracks rotational direction fluctuations to correct X-ray source errors and detector pile-up artifacts.

Processing circuitry applies machine learning to medical data and numerical acquisition conditions for high-fidelity image output.

MaxGIRF method applies analytic concomitant field models to MRI image reconstruction for improved signal fidelity.

Fourier ptychographic tomography reconstructs three-dimensional data from intensity measurements.

A computer-implemented method determines a final photon-count to penetrated depth function using test reconstructions and non-linear optimization.

Simulating angle-variable system blur during iterative reconstruction improves spatial resolution while suppressing noise amplification.

A deep learning energy transformation model generates dual-energy images from single-energy CT scans by selecting models based on contrast phase.

Time-domain system matrix eliminates point spread function artifacts in MPI reconstruction through velocity compensation and algebraic iteration.

An image processing apparatus derives common areas across projection images to emphasize effective regions in tomographic displays.

Evaluating operating current via signal-to-noise ratio eliminates external equipment complexity while maintaining measurement accuracy.

Virtual pulse injection logic probes discriminator state to measure count loss, correcting dead time errors that reduce nuclear imaging accuracy.

Segments large X-ray source focus into virtual sub-foci via computational processing to resolve high-energy industrial CT spatial resolution limits.

Image processing apparatus generates a gradient vector map to identify linear structural object directions and extrapolates missing side sections.