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

Artificial intelligence in fracture detection: transfer learning from deep convolutional neural networks.

D H Kim1, T MacKinnon1

  • 1Medical Imaging Department, Royal Devon and Exeter Hospital, Barrack Road, Exeter EX2 5DW, UK.

Clinical Radiology
|December 23, 2017
PubMed
Summary

Transfer learning using deep convolutional neural networks (CNNs) shows promise for automated fracture detection in radiographs. This approach achieved high accuracy, demonstrating potential for improved medical imaging workflows.

Related Concept Videos

Fractures: Bone Repair01:27

Fractures: Bone Repair

5.7K
Treatment for a fracture is based on the type of break, the bone affected, and the patient's age.
Minor fractures with no bone displacement are treated by immobilizing the fractured bone using a cast or splint. However, in the case of fractures with displaced bones, the broken bones are repositioned before immobilization to ensure successful healing without deformation and loss of function. The realignment of fractured bone ends is performed through a process called reduction. If the...
5.7K

You might also read

Related Articles

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

Sort by
Same author

Seroprevalence of canine dirofilariosis in South Korea.

Veterinary parasitology·2003
Same author

Impact of transplanted CD34+ cell dose in allogeneic unmanipulated peripheral blood stem cell transplantation.

Bone marrow transplantation·2003
Same author

Expression of HMGI(Y) associated with malignant phenotype of human gastric tissue.

Histopathology·2003
Same author

Search for long-lived charged massive particles in pp collisions at square root s = 1.8 TeV.

Physical review letters·2003
Same author

Search for a W' boson decaying to a top and bottom quark pair in 1.8 TeV pp collisions.

Physical review letters·2003
Same author

Competitive adsorption for binary mixture of 4-nitrophenol and phenol on RSTA using GAC.

Water science and technology : a journal of the International Association on Water Pollution Research·2003

Area of Science:

  • Radiology
  • Artificial Intelligence
  • Medical Imaging

Background:

  • Deep convolutional neural networks (CNNs) show potential for image analysis.
  • Transfer learning allows models trained on one task to be adapted for another.
  • Automated fracture detection on plain radiographs is an area of active research.

Purpose of the Study:

  • To evaluate the effectiveness of transfer learning from CNNs pre-trained on non-medical images for automated fracture detection.
  • To develop a CNN model for classifying wrist radiographs as "fracture" or "no fracture".

Main Methods:

  • The Inception v3 CNN model's top layer was re-trained using lateral wrist radiographs.
  • Data augmentation was applied to an initial set of 1,389 radiographs (695 fracture, 694 no fracture).

Related Experiment Videos

  • The dataset was split into training (80%), validation (10%), and testing (10%) groups, with an additional 100 images for final validation.

    • Main Results:

    • The model achieved an area under the receiver operator characteristic curve (AUC) of 0.954.
    • Sensitivity and specificity were 0.9 and 0.88, respectively, at an optimized diagnostic threshold.
    • The results demonstrate high diagnostic performance for fracture detection.

    Conclusions:

    • Transfer learning from CNNs is a viable approach for automated fracture detection on plain radiographs.
    • This method achieved state-of-the-art performance with a moderate sample size.
    • The technique has broad applications in medical imaging, potentially enhancing workflow and reducing clinical risk.