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

Oxygen-Assisted MOCVD Growth of Monolayer PtSe<sub>2</sub> Films With Bandgap Opening for Semiconducting FET Channels.

Advanced science (Weinheim, Baden-Wurttemberg, Germany)·2026
Same author

Scan Path Optimization and YOLO-Based Detection for Defect Inspection of Curved and Glossy Surfaces.

Sensors (Basel, Switzerland)·2026
Same author

Geometric morphometrics based diagnostic model for Skeletal Class III patients.

Communications medicine·2026
Same author

AI-validated fusion proteins for local inhibition of interleukin-17A.

Journal of controlled release : official journal of the Controlled Release Society·2026
Same author

Early hemoglobin levels after kidney transplantation predict clinical outcomes.

Kidney research and clinical practice·2026
Same author

Additive effects of ceftiofur-neomycin combination against multidrug-resistant pathogenic <i>Escherichia coli</i> in a murine infection model.

Veterinarni medicina·2026

Related Experiment Video

Updated: Nov 23, 2025

Deep Neural Networks for Image-Based Dietary Assessment
13:19

Deep Neural Networks for Image-Based Dietary Assessment

Published on: March 13, 2021

9.7K

How much deep learning is enough for automatic identification to be reliable?

Jun-Ho Moon, Hye-Won Hwang, Youngsung Yu

    The Angle Orthodontist
    |December 30, 2020
    PubMed
    Summary
    This summary is machine-generated.

    Developing accurate artificial intelligence (AI) for cephalometric landmark identification requires a substantial amount of learning data. At least 2300 labeled cephalograms are needed to achieve human-level accuracy in AI models.

    Keywords:
    Artificial intelligenceData quantityDeep learningLogarithmic transformation

    More Related Videos

    Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images
    08:20

    Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images

    Published on: October 27, 2023

    2.1K
    Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
    03:31

    Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

    Published on: December 15, 2023

    814

    Related Experiment Videos

    Last Updated: Nov 23, 2025

    Deep Neural Networks for Image-Based Dietary Assessment
    13:19

    Deep Neural Networks for Image-Based Dietary Assessment

    Published on: March 13, 2021

    9.7K
    Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images
    08:20

    Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images

    Published on: October 27, 2023

    2.1K
    Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
    03:31

    Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

    Published on: December 15, 2023

    814

    Area of Science:

    • Medical Imaging
    • Artificial Intelligence
    • Orthodontics

    Background:

    • Cephalometric landmark identification is crucial in orthodontic diagnosis and treatment planning.
    • Automating this process with artificial intelligence (AI) can improve efficiency and consistency.
    • Determining the necessary volume of training data for AI development is a key challenge.

    Purpose of the Study:

    • To ascertain the optimal quantity of learning data required for developing AI capable of automatic cephalometric landmark identification.
    • To establish a predictive model for estimating the data requirements for AI training.

    Main Methods:

    • Trained 96 artificial intelligence (AI) models using 2200 cephalograms with 80 manually identified landmarks.
    • Varied the quantity of learning data (50-2000 images) and the number of detection targets (19, 40, 80) per image.
    • Evaluated AI accuracy using radial error on a separate test set of 200 images.

    Main Results:

    • AI accuracy demonstrated a linear increase with more learning data on a logarithmic scale.
    • Accuracy decreased as the number of detection targets per image increased.
    • A prediction model indicated that at least 2300 learning data sets are necessary for human-level AI accuracy.

    Conclusions:

    • Developing accurate AI for cephalometric analysis necessitates a significant volume of training data.
    • The findings provide a foundational understanding for determining data requirements in AI development for medical imaging applications.