Detection of air bubbles in images of samples in well plates

A CNN-based method for bubble detection in well plate images addresses the challenge of bubble interference in light scattering measurements, enhancing measurement accuracy and sensitivity by identifying and mitigating bubble noise.

JP2026520325APending Publication Date: 2026-06-23WYATT TECHNOLOGY CORP
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
JP · JP
Patent Type
Applications
Current Assignee / Owner
WYATT TECHNOLOGY CORP
Filing Date
2023-07-27
Publication Date
2026-06-23

AI Technical Summary

Technical Problem

Current technologies face challenges in accurately detecting air bubbles in images of samples in well plates, which can lead to inaccurate light scattering measurements due to background noise, affecting sensitivity and requiring higher sample concentrations to compensate for bubble interference.

Method used

A computer-aided method using a convolutional neural network (CNN) trained via stochastic gradient descent is employed to analyze images of well plates, identifying the presence of bubbles by processing cropped images and providing probability values for bubble detection.

Benefits of technology

Enhances the accuracy of light scattering measurements by effectively detecting and removing bubble interference, improving sensitivity and reducing the need for higher sample concentrations.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 2026520325000001_ABST
    Figure 2026520325000001_ABST
Patent Text Reader

Abstract

This disclosure describes apparatus (612), method (200), system (600), and computer program product (642) for detecting bubbles in images of a sample in the wells of a well plate. In exemplary embodiments, the computer implementation, system, and computer program product include: a computer system (600) receiving at least one image of at least one well of at least one well of a well plate containing a liquid sample (210); the computer system (600) performing a set of logical operations to crop the image to include the well of interest (212); and the computer system (600) performing a set of logical operations to process the cropped image using a trained convolutional neural network to obtain a probability that the image depicts at least one bubble (214).
Need to check novelty before this filing date? Find Prior Art