Reconstruction From Projection TRIZ Cases

SWIM extracts phase information from a single scan to generate orientation-independent susceptibility maps, eliminating the need for multiple scans.

Modulating PET image reconstruction smoothing parameters via data-independent sensitivity factors to achieve uniform spatial resolution.

An MRI system generates additional k-space data to maintain signal intensity continuity during image reconstruction.

Dynamic reconstruction using GPU backprojection eliminates tissue superimposition blurring in mammography by enabling real-time on-demand 3D image generation.

Segment cardiac images by phase to reduce motion blur and improve 3D reconstruction accuracy.

Segmented axial reconstruction reduces memory requirements and statistical noise while maintaining high image quality in whole-body PET scans.

Comparison algorithms match mask images to patient movement states in digital subtraction angiography, reducing motion artifacts without manual allocation.

Multi-energy computed tomography differentiates materials with similar attenuation by analyzing weighted Hounsfield unit values across varying energy levels.

Computing a local penalty function reduces its value in overlap regions to prevent over-smoothing and maintain clinically relevant features.

Infrared spectrometer maps gas distribution while maintaining operator safety.

Dynamic gate numbers adapt to voxel motion ranges during binning, resolving respiratory blur in chest examinations while maintaining computational efficiency.

Voxel-based volumetric modeling reconstructs 3D marker shapes from binary projections, resolving segmentation reliability issues caused by varying orientations.

Quasi-periodic gratings modulate X-ray phase to enhance contrast resolution in soft tissue imaging.

Segmenting the system matrix into a geometrical projection and an image blurring component reduces computational overhead while maintaining high reconstruction accuracy.

A computed tomography system uses focal spot flying and rapid kV switching to acquire multi-energy data through a single detector pixel.

Augmented reality overlay renders partial medical image data based on endoscope position, eliminating monitor glances to reduce surgery time.

Guidewire tracking registers position within a body cavity to reassign voxel radiodensity values, resolving mucus interference that reduces image contrast.

Navigation system calculates screw trajectory and entry point metrics to reduce pull-out risk during spinal stabilization.

Real-time resampling converts non-voxel elements to voxels, preserving image quality and reducing interpolation artifacts.

Maximum correlated kurtosis deconvolution filters microwave brain signals to isolate target responses for bleeding detection.

A medical image processing apparatus generates motion information using partial angle reconstruction image pairs to produce reconstructed images.

A three-dimensional lookup table assigns monochromatic attenuation values to polychromatic data for CT scanners.

An adaptive image processing method reduces false and missing parts in reconstructed images by iteratively selecting tilt angles that fill information gaps.

Recursive backprojection with coordinate transformations reduces computational bottlenecks in high-resolution tomographic reconstruction.

A medical image processing apparatus uses a derivative of planar integrals motion index to select consistent projection data for reconstruction.

Invertible wavelet transforms enable exact interior reconstruction from truncated projections, resolving non-uniqueness issues in compressive sensing.

Shift-variant weighting corrects data redundancy during variable speed helical scanning, resolving slice-view center mismatches that generate image artifacts.

A system dynamically reconstructs medical image montages to match specific display device characteristics.

A hybrid approach selects subsets of large tomographic datasets in frequency space while maintaining others in real space.

A processor dynamically assembles reconstruction flows from algorithm containers to process raw data into target images.

Separating static and dynamic components in multi-dimensional image data improves time resolution without requiring extensive projection image acquisition.

Segmented detector arms retract into the bore to improve patient comfort while maintaining imaging accuracy through dynamic positioning.

Pre-associated lookup tables define optimal region parameters for each mode, eliminating manual selection delays while maintaining measurement precision.

A neural network modeled on the Cooley-Tukey fast Fourier transform learns optimal reconstruction operators for medical imaging systems.

Adaptive sparsifying transforms learn from data to reconstruct high-quality images from undersampled measurements, reducing artifacts and radiation exposure.

An out-of-distribution testing neural network analyzes reconstructed magnetic resonance images to detect noise and artifacts before full processing.

Prior image constrained compressed sensing reconstructs high signal-to-noise ratio images using sparsified data and Lagrange multipliers.

Cascade modules and a selection unit reduce latency while maintaining high spatial resolution in cardiac imaging.

Resampling Radon space data via machine learning corrects beam hardening artifacts, improving CT image quality and quantification accuracy.

Segmented sensor plates enable adaptive electrical capacitance volume tomography to resolve center region details despite signal-to-noise constraints.

A data processing system organizes ultrasonic transit data into matrices for face recognition and Bayesian classification to identify multiphase flow regimes.

Generates synthesized projections from initial image data to refine dual-energy computed tomography reconstruction without requiring rapid generator switching.

A Wasserstein GAN with a generator network learns filtered back-projection to reduce streak artifacts from under-sampled data in few-view CT systems.

Extract high-frequency components from high kVp projections and add them to low kVp data to reduce noise while maintaining spatial resolution.

A cardiac phase determination method calculates regional motion variations to identify the optimal imaging window for X-ray CT scans.

Poly-energetic reconstruction addresses photon starvation and beam hardening by applying weighted iterative updates to volume estimates.

A linear correction model optimizes parameters using epipolar geometry consistency to reduce X-ray image artifacts.

Estimates missing low energy projection data from high energy beams to resolve undersampling artifacts, improving signal-to-noise ratio in reconstructed images.

A depth map links two-dimensional radiological image positions to tomographic plane coordinates.

A video processing apparatus clips and combines partial video data from multiple cameras to generate customizable synthesized patterns.