Microscope Sample Localization Using Deep Learning Stage Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current charged particle microscope systems require extensive manual effort for sample identification and tracking, limiting technician productivity and introducing inefficiencies in the microscopy process.
Innovation Solution
The implementation of a scientific instrument support system that utilizes a navigation camera, trained models for image analysis, and automated stage coordinate association to identify samples and determine their locations on a fixture, thereby automating the sample identification and tracking process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If manual sample identification and tracking is used, then technicians can handle samples, but productivity is limited and time is consumed
Solution Approach 1:
The system enables self-service automation where the microscope system automatically identifies samples, tracks their locations, and manages stage coordinates without requiring manual technician intervention. The navigation camera captures images, the trained model analyzes them to identify samples and determine positions, and the system automatically associates stage coordinates, allowing technicians to focus on higher-value tasks while the system handles routine identification and tracking operations autonomously
Solution Approach 2:
The patent replaces manual mechanical operations with an automated vision-based system. Instead of technicians manually identifying and tracking samples, a navigation camera captures images, a trained deep learning model analyzes the images to identify samples and determine their locations, and the system automatically associates stage coordinates. This substitution of manual mechanical processes with automated optical and computational processes significantly increases productivity and reduces time consumption
2Productivity
If automated sample identification is implemented, then throughput increases, but system complexity increases
Solution Approach 1:
The navigation camera serves multiple functions: it captures images of samples on the fixture, provides visual feedback for sample location, and enables the trained model to identify samples and determine their positions. This multi-functional component reduces the need for separate specialized devices, thereby increasing throughput while limiting the increase in system complexity
Solution Approach 2:
The system creates a digital representation (image) of the physical sample arrangement on the fixture. The navigation camera captures this visual copy, which is then processed by the trained model to identify samples and determine their locations without requiring physical manipulation or complex mechanical tracking systems. This copying approach simplifies the automation system while maintaining high throughput
Data Source
AI summary
Disclosed herein are scientific instrument support systems, as well as related methods, computing devices, and computer-readable media. For example, in some embodiments, a method for determining sample location and associated stage coordinates by a microscope at least comprises acquiring, with a navigation camera, an image of a plurality of samples loaded on a fixture, the image being of low resolution at a field of view that includes the fixture and all samples of the plurality of samples, analyzing the image with a trained model to identify the plurality of samples, based on the analysis, associating each sample with a location on the fixture, based on the location on the fixture of each sample, associating separate stage coordinate information with each sample of the plurality of samples loaded on the fixture, and translating a stage holding the fixture to first stage coordinates based on the associated stage coordinate information of a first sample of the plurality of samples.