Detection method, detection system, noise removal model training method, and noise removal model training system

The denoising model training method and system address the challenge of noise interference in detection systems by using neural networks and autoencoders to remove noise from detection signals, enhancing the accuracy and sensitivity of target detection in flow cytometry and similar applications.

WO2024228394A1PCT designated stage expired Publication Date: 2024-11-07THE UNIV OF TOKYO +1

Patent Information

Application Number
PCT/JP2024/016828
Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Priority Date
2023-12-01
Filing Date
2024-05-01
Publication Date
2024-11-07

AI Technical Summary

Technical Problem

Current detection systems face challenges in accurately detecting targets due to noise interference, particularly in flow cytometry and other detection devices, where noise removal methods are inadequate, leading to suboptimal signal processing and target identification.

Method used

A denoising model training method and system that utilizes a noise removal model, trained using noise and learning signals, to effectively remove noise from detection signals, enabling accurate target detection by employing neural networks and autoencoders, such as variational autoencoders, to process one-dimensional to high-dimensional signals.

Benefits of technology

This approach significantly enhances the accuracy of target detection by effectively removing noise, allowing for the identification of targets with high sensitivity, even at low concentrations or small particle sizes, without the need for pre-treatment processes like ultracentrifugation, and improves the detection of biomolecules and extracellular vesicles.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure JP2024016828_07112024_PF_FP_ABST
    Figure JP2024016828_07112024_PF_FP_ABST
Patent Text Reader

Abstract

This detection system comprises: a detection signal acquisition unit 306 that acquires, as a detection signal, a signal from an object being detected; and a noise-removed signal acquisition unit 307 that inputs the detection signal to a noise removal model 304 trained using noise signals obtained when no object is detected, and acquires at least one noise-removed signal, from which noise has been removed.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Method for measuring nucleated red blood cells and white blood cells in whole blood samples using an automated blood analyzer.

    JP5230648B2

  • Method and system for sample analysis

    JP7134184B2

  • Signal processing method, signal processing device, and signal processing system

    JP7329165B1

  • Methods and systems for cytometry

    JP7369385B2

  • Shape parameters for hematological instruments

    JP2012505470A

Cited By

  • Self-referencing spectral constraint-based spectral reconstruction method and apparatus, device, and medium

    US12742679B2

  • Microparticle analysis system, microparticle analysis device, and microparticle analysis method

    WO2026204449A1