Jove
Visualize
Contact Us
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

Related Experiment Videos

Data processing for multi-channel optical recording: action potential detection by neural network.

S Yamada1, H Kage, M Nakashima

  • 1Biotechnology Department, Mitsubishi Electric Corporation, Hyogo, Japan.

Journal of Neuroscience Methods
|June 1, 1992
PubMed
Summary
This summary is machine-generated.

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

Prediction model for ctDNA detectability in liquid comprehensive genomic profiling of advanced pancreatic cancer based on Japanese real-world data.

ESMO open·2026
Same author

Pan-cancer clinical and molecular landscape of MTAP deletion in nationwide and international comprehensive genomic data.

ESMO open·2025
Same author

A machine learning-based analysis of nationwide cancer comprehensive genomic profiling data across cancer types to identify features associated with recommendation of genome-matched therapy.

ESMO open·2024
Same author

A second-generation profiling system for quantitative methylation analysis of multiple gene promoters: application to lung cancer.

Oncogene·2007
Same author

[Psychosocial approach of the epileptic patient: analysis of 30 adults].

Arquivos de neuro-psiquiatria·1992
Same author

Manganese-containing superoxide dismutase in blood and urine during open-heart surgery.

Japanese circulation journal·1992
Same journal

Detection of cochlear microphonic for differential diagnosis between auditory neuropathy mice and noise-induced sensorineural hearing loss mice.

Journal of neuroscience methods·2026
Same journal

Assessment metrics for pain control in rats: A methodological commentary.

Journal of neuroscience methods·2026
Same journal

Infant EEG preprocessing pipelines: A capability framework and current gaps in practice.

Journal of neuroscience methods·2026
Same journal

Methods for measuring neural activity during voluntary wheel running.

Journal of neuroscience methods·2026
Same journal

Serotype-dependent differences in AAV cellular transduction rates in the hypothalamus of Arctic ground squirrels.

Journal of neuroscience methods·2026
Same journal

Rapid generation of human sensory neurons from iPSC for modeling of peripheral neuropathies.

Journal of neuroscience methods·2026
See all related articles

A new neural network program quickly and accurately detects action potentials (APs) in optical recordings. Improved classification with signal-to-noise ratio achieved 96% detection accuracy, enabling mostly automatic AP identification.

Area of Science:

  • Neuroscience
  • Computational Biology
  • Signal Processing

Background:

  • Accurate detection of action potentials (APs) is crucial for analyzing neural activity from optical recordings.
  • Manual analysis of large electrophysiological datasets is time-consuming and prone to errors.

Purpose of the Study:

  • To develop a fast and precise program for automatic action potential detection using neural networks.
  • To improve the accuracy of AP detection in raw multi-channel optical recording data.

Main Methods:

  • A two-step approach was employed: peak detection in raw optical data followed by neural network classification.
  • The neural network was trained using the backpropagation algorithm with thousands of manually classified peaks.
  • Signal-to-noise ratio (SNR) was incorporated to enhance classification performance.

Related Experiment Videos

Main Results:

  • The neural network classifier outperformed template matching and nearest-neighbor methods.
  • Incorporating SNR improved classification accuracy, achieving 96% detection of manually classified APs.
  • The developed program allows for mostly automatic AP detection with minimal human intervention.

Conclusions:

  • The neural network-based program offers a significant advancement in automated action potential detection.
  • The method provides a fast, precise, and largely automated solution for analyzing optical recording data.
  • Further improvements may reduce the need for human intervention in classifying undecided peaks.