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

Patient-Facing Radiology Communication with LLMs: Calibration Deficit and the Metadata Paradox.

Healthcare (Basel, Switzerland)·2026
Same author

Strategies to Increase Patients' Adherence to Digital Therapeutics for Musculoskeletal Diseases: A Narrative Review.

Yonsei medical journal·2026
Same author

Evaluation of a newly developed hybrid brace for scoliosis: Usability tests and subject-specific finite element analysis.

Technology and health care : official journal of the European Society for Engineering and Medicine·2025
Same author

Predictive Factors for Postoperative Outcomes of Cervical Spondylotic Myelopathy in Individuals With Cerebral Palsy.

Global spine journal·2025
Same author

A deep-learning algorithm using real-time collected intraoperative vital sign signals for predicting acute kidney injury after major non-cardiac surgeries: A modelling study.

PLoS medicine·2025
Same author

Salivary Gland Volume Changes and Dry Mouth Symptom Following Definitive Radiation Therapy in Oropharyngeal Cancer Patients-A Comparison of Two Different Approaches: Intensity-Modulated Radiation Therapy Versus Intensity-Modulated Radiation Therapy/Intensity-Modulated Proton Therapy Combination.

Cancers·2025
Same journal

The Role of YouTube as an Information Platform for Ovarian Cancer: the Educational Quality and Reliability.

Yonsei medical journal·2026
Same journal

Bone Bridge Effect for the Treatment of Acute Osteoporotic Vertebral Compression Fractures: A Multistrategic Approach Using an Anabolic Agent.

Yonsei medical journal·2026
Same journal

Outcomes and Prognostic Factors in Hepatopancreatoduodenectomy.

Yonsei medical journal·2026
Same journal

The Risk of Incident Diabetes Mellitus in Relation to Egg Consumption among Working-Aged Korean Adults.

Yonsei medical journal·2026
Same journal

SpyGlass Direct Visualization System: A Cost-Effective Approach Enhancing Nutritional and Immune Recovery in Difficult Bile Duct Stone Management.

Yonsei medical journal·2026
Same journal

Optimizing Antibiotic Use for Urinary Tract Infections: A Qualitative Assessment of Regular Prescribing Practices.

Yonsei medical journal·2026
See all related articles

Related Experiment Video

Updated: Sep 14, 2025

In Vivo Quantification of Hip Arthrokinematics during Dynamic Weight-bearing Activities using Dual Fluoroscopy
07:43

In Vivo Quantification of Hip Arthrokinematics during Dynamic Weight-bearing Activities using Dual Fluoroscopy

Published on: July 2, 2021

3.2K

Deep Learning-Based Landmark Detection Model for Multiple Foot Deformity Classification: A Dual-Center Study.

Su Ji Lee1, Hangyul Yoon2, Seongsu Bae2

  • 1Department and Research Institute of Rehabilitation Medicine, Yonsei University College of Medicine, Seoul, Korea.

Yonsei Medical Journal
|July 25, 2025
PubMed
Summary
This summary is machine-generated.

A new heatmap-in-heatmap (HIH) model automates foot deformity diagnosis from radiographs, improving accuracy and efficiency over manual methods. This AI tool offers a reliable solution for clinical use.

Keywords:
Artificial intelligencedeep learningdiagnostic imagingfoot deformities

More Related Videos

Author Spotlight: Three-Dimensional Cephalometric Landmark Annotation Demonstration on Human Cone Beam Computed Tomography Scans
10:23

Author Spotlight: Three-Dimensional Cephalometric Landmark Annotation Demonstration on Human Cone Beam Computed Tomography Scans

Published on: September 8, 2023

3.0K
High-Resolution Three-Dimensional Imaging of the Footpad Vasculature in a Murine Hindlimb Gangrene Model
08:16

High-Resolution Three-Dimensional Imaging of the Footpad Vasculature in a Murine Hindlimb Gangrene Model

Published on: March 16, 2022

3.6K

Related Experiment Videos

Last Updated: Sep 14, 2025

In Vivo Quantification of Hip Arthrokinematics during Dynamic Weight-bearing Activities using Dual Fluoroscopy
07:43

In Vivo Quantification of Hip Arthrokinematics during Dynamic Weight-bearing Activities using Dual Fluoroscopy

Published on: July 2, 2021

3.2K
Author Spotlight: Three-Dimensional Cephalometric Landmark Annotation Demonstration on Human Cone Beam Computed Tomography Scans
10:23

Author Spotlight: Three-Dimensional Cephalometric Landmark Annotation Demonstration on Human Cone Beam Computed Tomography Scans

Published on: September 8, 2023

3.0K
High-Resolution Three-Dimensional Imaging of the Footpad Vasculature in a Murine Hindlimb Gangrene Model
08:16

High-Resolution Three-Dimensional Imaging of the Footpad Vasculature in a Murine Hindlimb Gangrene Model

Published on: March 16, 2022

3.6K

Area of Science:

  • Medical imaging analysis
  • Artificial intelligence in healthcare
  • Orthopedic diagnostics

Background:

  • Manual diagnosis of foot deformities from radiographs is labor-intensive and prone to variability.
  • Automated diagnostic tools are needed to improve efficiency and consistency.

Purpose of the Study:

  • To introduce a novel heatmap-in-heatmap (HIH)-based model for automated diagnosis of foot deformities.
  • To evaluate the HIH model's performance using weight-bearing foot radiographs.

Main Methods:

  • A dual-center retrospective study involving 1561 (training) and 374 (validation) images from the first center, and 527 images for external validation from the second center.
  • Diagnosis of five foot deformities using angles between predicted landmarks in anterior-posterior (AP) and lateral radiographs.
  • Comparison with a baseline model, FlatNet.

Main Results:

  • The HIH model achieved higher accuracy (85.1% vs. 78.9%), sensitivity (84.1% vs. 78.9%), and specificity (85.9% vs. 79.0%) compared to FlatNet.
  • HIH demonstrated lower error rates, higher detection rates, faster processing speeds, and fewer parameters.
  • Robust performance was confirmed in both internal and external validation sets.

Conclusions:

  • The heatmap-in-heatmap (HIH) model shows high efficacy for automated diagnosis of multiple foot deformities.
  • This AI-driven approach is promising for various landmark-based medical imaging tasks.