US 12,718,320 B2Granted

X-ray super-resolution assessment via spatial filtering

AssigneeCarl Zeiss X-ray Microscopy, Inc.

InventorMatthew Andrew

Life of this patent

Estimating the term…

Loading patent…

Abstract

A technique is disclosed for analyzing and displaying the extent to which the images and structures inferred by a physically seeded multiscale network correspond to genuine resolution improvement and the extent to which they correspond to the hallucination of realistic looking structures. A selected reconstruction generated using a trained neural network is compared against a baseline representation (e.g., a baseline reconstruction) by calculating image similarity metrics between progressing spatially filtered versions of selected reconstruction. As the amount of spatial filtering increases, at some point the image similarity metrics will reach an extrema (e.g., a lowest distance between the images). At that extrema, the parameter(s) of the spatial filter can be used to identify a resolution score (e.g., a length scale) associated with that extrema. The resolution score is indicative of an amount of resolution recovery associated with the trained neural network with respect to the baseline.

Claims20

Dependent claims

A method, comprising:

receiving an improved reconstruction of a subject, the improved reconstruction generated by supplying a trained neural network with first imaging data acquired of the subject, the first imaging data acquired using an electromagnetic radiation imager;
receiving a baseline representation of the subject;
calculating a plurality of image similarity metrics between the baseline representation and the improved reconstruction, wherein, for each of the plurality of image similarity metrics, calculating the plurality of image similarity metrics includes:
applying a spatial filter to the improved reconstruction based at least in part on a filter parameter;
calculating an image similarity metric between the baseline representation and the filtered improved reconstruction;
and
adjusting the filter parameter;
determining a selected filter parameter associated with an extrema image similarity metric of the plurality of image similarity metrics representative of a closest similarity between the baseline representation and the filtered improved reconstruction;
and
generating a resolution score associated with the selected filter parameter.

7 more elements

Dependencies are inferred from the claim text.

Facts

Application: 18/680,854Filed 31 May 2024
Published: US 2025/0371656 A14 Dec 2025
Granted: 25 Aug 2026
Term ends: Not estimatedThe source records don’t support an estimate.

Classification12 codes

G06T 3/4046
Scaling of whole images or parts thereof, e.g. expanding or contracting: using neural networks
G06T 12/30
Tomographic reconstruction from projections: image post-processing, e.g. metal artefact correction
G06T 5/20
Image enhancement or restoration: using local operators
G06T 5/60
Image enhancement or restoration: using machine learning, e.g. neural networks