Jove
Visualize
Contact Us

Related Experiment Video

Updated: Sep 25, 2025

Microfluidic Chips Controlled with Elastomeric Microvalve Arrays
18:11

Microfluidic Chips Controlled with Elastomeric Microvalve Arrays

Published on: October 1, 2007

21.3K

Microfluidic Devices Controlled by Machine Learning with Failure Experiments.

Kenta Fukada1, Michiko Seyama1

  • 1NTT Device Technology Laboratories, NTT Corporation, 3-1 Morinosato, Wakamiya, Atsugi, Kanagawa 243-0198, Japan.

Analytical Chemistry
|April 25, 2022
PubMed
Summary

This study introduces reinforced learning for microfluidic particle sorting, using failed experiments to improve training efficiency. This approach accelerates the optimization of microchannel devices for isolating cells and other biological samples.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Thermally Degradable Water Diffusion Barrier Assembled by Gelatin and Beeswax toward Edible Electronics.

ACS applied materials & interfaces·2024
Same author

Noninvasive Glucose Sensing in Dielectrically Equivalent Multilayer Skin Phantoms.

Langmuir : the ACS journal of surfaces and colloids·2023
Same author

Thermally Degradable Inductors with Water-Resistant Metal Leaf/Oleogel Wires and Gelatin/Chitosan Hydrogel Films.

ACS applied materials & interfaces·2022
Same author

Thermoresponsive Gelatin/Chitosan Hydrogel Films for a Degradable Capacitor.

ACS applied materials & interfaces·2021
Same author

Direct Measurement of Near-Wall Molecular Transport Rate in a Microchannel and Its Dependence on Diffusivity.

Langmuir : the ACS journal of surfaces and colloids·2021
Same author

Trace Material Capture by Controlled Liquid Droplets on a Superhydrophobic/Hydrophilic Surface.

Analytical chemistry·2017
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Area of Science:

  • Microfluidics and Cell Sorting
  • Machine Learning in Biological Engineering
  • Robotic Automation for Scientific Research

Background:

  • Microchannel techniques are crucial for isolating specific objects like cells in biological solutions.
  • Traditional particle sorting in microfluidic devices is often time-consuming and labor-intensive due to unpredictable particle behavior.

Purpose of the Study:

  • To develop a more efficient method for microfluidic particle sorting using reinforced learning.
  • To maximize the training effect with limited data by utilizing failure results in the learning process.

Main Methods:

  • Employed reinforced learning with microscopic images of the microfluidic separation process, including failed conditions.
  • Implemented gradient rewards based on the degree of failure (e.g., inappropriate flow speeds, dilution rates) to guide learning.

More Related Videos

Generation of Dynamical Environmental Conditions using a High-Throughput Microfluidic Device
14:48

Generation of Dynamical Environmental Conditions using a High-Throughput Microfluidic Device

Published on: April 17, 2021

4.2K
A Microfluidic Device for Studying Multiple Distinct Strains
08:15

A Microfluidic Device for Studying Multiple Distinct Strains

Published on: November 9, 2012

8.9K

Related Experiment Videos

Last Updated: Sep 25, 2025

Microfluidic Chips Controlled with Elastomeric Microvalve Arrays
18:11

Microfluidic Chips Controlled with Elastomeric Microvalve Arrays

Published on: October 1, 2007

21.3K
Generation of Dynamical Environmental Conditions using a High-Throughput Microfluidic Device
14:48

Generation of Dynamical Environmental Conditions using a High-Throughput Microfluidic Device

Published on: April 17, 2021

4.2K
A Microfluidic Device for Studying Multiple Distinct Strains
08:15

A Microfluidic Device for Studying Multiple Distinct Strains

Published on: November 9, 2012

8.9K
  • Utilized failure data to enhance training datasets and accelerate convergence to optimal solutions.

    • Main Results:

    • Demonstrated that reinforced learning, incorporating failure analysis, significantly improves training efficiency for microfluidic control.
    • Showcased the ability to automatically find optimal separation conditions for new, related samples after initial learning.
    • Established that failed experiments contribute valuable data, reducing wasted resources and improving accuracy.

    Conclusions:

    • Microfluidic control enhanced by reinforced learning, which leverages failure data, offers a powerful approach to particle sorting.
    • This method accelerates the optimization of microfluidic devices, making them more efficient and less labor-intensive.
    • The developed device control strategy has broad applicability in automatic synthetic chemistry, biomedical analysis, and microfabrication robotics.