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 Concept Videos

You might also read

Related Articles

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

Sort by
Same author

An Empirical Bayes approach for the study of phenotypic evolution from high-dimensional data.

Systematic biology·2026
Same author

Detecting environmentally dependent developmental plasticity in fossilized individuals.

Proceedings of the National Academy of Sciences of the United States of America·2025
Same author

Divergent patterns of cranial suture fusion in marsupial and placental mammals.

Zoological journal of the Linnean Society·2025
Same author

Exploring fluctuating asymmetry in two recovering populations of the Eurasian otter.

Biology letters·2024
Same author

Conservation implications of a mismatch between data availability and demographic impact.

Ecology and evolution·2023
Same author

High-Density Geometric Morphometric Analysis of Intraspecific Cranial Integration in the Barred Grass Snake (<i>Natrix helvetica</i>) and Green Anole (<i>Anolis carolinensis</i>).

Integrative organismal biology (Oxford, England)·2023

Related Experiment Video

Updated: Jun 18, 2025

From Voxels to Knowledge: A Practical Guide to the Segmentation of Complex Electron Microscopy 3D-Data
12:08

From Voxels to Knowledge: A Practical Guide to the Segmentation of Complex Electron Microscopy 3D-Data

Published on: August 13, 2014

24.5K

How many specimens make a sufficient training set for automated three-dimensional feature extraction?

James M Mulqueeney1,2, Alex Searle-Barnes1, Anieke Brombacher1

  • 1School of Ocean & Earth Science, National Oceanography Centre Southampton, University of Southampton Waterfront Campus, Southampton, UK.

Royal Society Open Science
|August 5, 2024
PubMed
Summary

Deep learning models require at least eight training images for 95% accuracy in segmenting planktonic foraminifera. Data augmentation can improve accuracy by 8.0% for 3D image analysis.

Keywords:
data augmentationdeep learningfeature extractionimage segmentationplanktonic foraminifera

More Related Videos

Label-Free Identification of Lymphocyte Subtypes Using Three-Dimensional Quantitative Phase Imaging and Machine Learning
08:58

Label-Free Identification of Lymphocyte Subtypes Using Three-Dimensional Quantitative Phase Imaging and Machine Learning

Published on: November 19, 2018

12.5K
Author Spotlight: Advanced Techniques for Characterizing Tissue Mineralization in Bone Regeneration Research
07:29

Author Spotlight: Advanced Techniques for Characterizing Tissue Mineralization in Bone Regeneration Research

Published on: September 27, 2024

695

Related Experiment Videos

Last Updated: Jun 18, 2025

From Voxels to Knowledge: A Practical Guide to the Segmentation of Complex Electron Microscopy 3D-Data
12:08

From Voxels to Knowledge: A Practical Guide to the Segmentation of Complex Electron Microscopy 3D-Data

Published on: August 13, 2014

24.5K
Label-Free Identification of Lymphocyte Subtypes Using Three-Dimensional Quantitative Phase Imaging and Machine Learning
08:58

Label-Free Identification of Lymphocyte Subtypes Using Three-Dimensional Quantitative Phase Imaging and Machine Learning

Published on: November 19, 2018

12.5K
Author Spotlight: Advanced Techniques for Characterizing Tissue Mineralization in Bone Regeneration Research
07:29

Author Spotlight: Advanced Techniques for Characterizing Tissue Mineralization in Bone Regeneration Research

Published on: September 27, 2024

695

Area of Science:

  • Paleontology
  • Computational Biology
  • Image Analysis

Background:

  • Manual image segmentation is time-consuming and prone to bias.
  • Deep learning offers automated feature extraction from 3D images.
  • Optimal training data size and data augmentation effects are under-explored.

Purpose of the Study:

  • Determine minimum training set size for accurate 3D image segmentation.
  • Assess data augmentation's impact on accuracy and efficiency.
  • Compare segmentation challenges for internal vs. external structures.

Main Methods:

  • Manual segmentation of 50 planktonic foraminifera (genus Menardella).
  • Training deep learning models with varying numbers of specimens.
  • Utilizing data augmentation to expand training sets.
  • Evaluating accuracy for volumetric and shape data extraction.

Main Results:

  • 95% accuracy achieved with eight training specimens.
  • Data augmentation improved network accuracy by up to 8.0%.
  • Internal structures are more challenging to segment than external ones due to low contrast and complexity.

Conclusions:

  • Provides insights into optimal training set sizes for 3D image segmentation.
  • Highlights data augmentation's potential for enhancing feature extraction.
  • Suggests deep learning models need careful consideration for complex internal structures.