Direct Electron Camera Imaging for Low-SNR Cluster Reconstruction

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing direct electron detection cameras face issues with non-uniform signal-to-noise ratios (SNRs) in clusters, leading to reduced SNR in regenerated images, especially during voltage reduction, affecting detective quantum efficiency and three-dimensional reconstruction resolution.

Innovation Solution

Classify clusters in original images into low and high SNR clusters, perform three-dimensional reconstruction separately, calculate a filtering function based on these models, and superpose filtered images to maximize SNR.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Quantity of substance

If the centroid method is used to process original images, then the number and size of output images are reduced, but the SNR of regenerated images decreases due to non-uniform SNR distribution in clusters

Engineering Contradiction:

Improvenumber of output imagesVSAvoidSNR of regenerated images

Core Design Contradiction:

Quantity of substanceVSMeasurement precision

Solution Approach 1:

The patent segments clusters into different SNR groups (low SNR clusters and high SNR clusters) based on their signal-to-noise ratio characteristics. This segmentation allows differential processing where low SNR clusters undergo filtering operations while high SNR clusters are processed differently, thereby preserving overall image quality while reducing output volume.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent applies different processing qualities to different regions/clusters based on their local SNR characteristics. Low SNR clusters receive filtering treatment to improve their quality, while high SNR clusters are processed with different parameters. This local quality adjustment ensures that each cluster is processed according to its specific needs, preventing uniform degradation of SNR across all clusters.

Inventive Principle:

Principle #3Local quality

2Device complexity

If low SNR clusters are directly used for image regeneration, then processing complexity is reduced, but noise increases and detective quantum efficiency decreases

Engineering Contradiction:

Improveprocessing complexityVSAvoiddetective quantum efficiency

Core Design Contradiction:

Device complexityVSMeasurement precision

Solution Approach 1:

The patent performs preliminary classification of clusters into different SNR groups before image regeneration. This preliminary action identifies low SNR clusters that require special filtering treatment, allowing the system to prepare appropriate processing strategies in advance rather than handling all clusters uniformly during the regeneration phase.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent changes processing parameters based on cluster SNR characteristics. Low SNR clusters are subjected to filtering operations with specific parameters designed to enhance their signal quality, while high SNR clusters use different parameters. This parameter adaptation optimizes the balance between processing complexity and output quality.

Inventive Principle:

Principle #35Parameter changes

3Speed

If the centroid method is used during voltage reduction, then processing speed is maintained, but backscattered electrons increase noise significantly

Engineering Contradiction:

Improveprocessing speedVSAvoidnoise from backscattered electrons

Core Design Contradiction:

SpeedVSObject-generated harmful factors

Solution Approach 1:

The patent converts the harmful effect of backscattered electrons into a useful classification criterion. By identifying clusters affected by backscattered electrons as low SNR clusters, the system applies targeted filtering to these specific regions. This transforms the previously harmful noise source into an opportunity for selective enhancement, maintaining processing speed while reducing the impact of backscattered electron noise.

Inventive Principle:

Principle #22Blessing in disguise (Convert harm into benefit)

Data Source

PatentUS12505977B2Imaging method and apparatus for direct electron detection cameras and computer device

Publication Date: 2025.12.23 INSTITUTE OF BIOPHYSICS CHINESE ACADEMY OF SCIENCES

AI summary

The present invention discloses an imaging method and an apparatus for direct electron detection cameras and a computer device, and relates to the technical field of electron microscope cameras. The present invention is mainly capable of improving the signal-to-noise ratio (SNR) of an image so as to improve the detective quantum efficiency of electron. The method includes the steps of classifying clusters in an original image to obtain low SNR clusters and high SNR clusters; performing three-dimensional reconstruction by using the images corresponding to the low SNR clusters and the high SNR clusters, respectively, to obtain three-dimensional models corresponding to the low SNR clusters and the high SNR clusters, respectively; performing filtering on the image corresponding to the low SNR clusters by using the filtering function, and superimposing the image to obtain m output image corresponding to the vitrified sample.