Jove
Visualize
Contact Us

Related Concept Videos

Classification of Systems-I01:26

Classification of Systems-I

215
Linearity is a system property characterized by a direct input-output relationship, combining homogeneity and additivity.
Homogeneity dictates that if an input x(t) is multiplied by a constant c, the output y(t) is multiplied by the same constant. Mathematically, this is expressed as:
215
Classification of Systems-II01:31

Classification of Systems-II

178
Continuous-time systems have continuous input and output signals, with time measured continuously. These systems are generally defined by differential or algebraic equations. For instance, in an RC circuit, the relationship between input and output voltage is expressed through a differential equation derived from Ohm's law and the capacitor relation,
178

You might also read

Related Articles

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

Sort by
Same journal

Delayed radial nerve palsy caused by elbow heterotopic ossification following a Monteggia Fracture.

Journal of Yeungnam medical science·2026
Same journal

Therapeutic efficacy of mesenchymal stem cells in periodontal regeneration: a systematic review and meta-analysis of clinical trials.

Journal of Yeungnam medical science·2026
Same journal

Forty-year evolution of research topics in the Journal of Yeungnam Medical Science: a text mining and keyword network analysis.

Journal of Yeungnam medical science·2026
Same journal

Are artificial intelligence (AI) agents ready for medicine and biomedical research? A narrative review.

Journal of Yeungnam medical science·2026
Same journal

Primary culture and characterization of human upper limb muscle satellite cells: an experimental study.

Journal of Yeungnam medical science·2026
Same journal

Fulminant streptococcal toxic shock syndrome presenting as cervical epidural abscess: a case report.

Journal of Yeungnam medical science·2026
See all related articles
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 Video

Updated: Jul 21, 2025

Author Spotlight: Advancing CBCT and Digital Dental Image Integration with AI-Assisted Digitization
05:49

Author Spotlight: Advancing CBCT and Digital Dental Image Integration with AI-Assisted Digitization

Published on: February 23, 2024

887

Classification of dental implant systems using cloud-based deep learning algorithm: an experimental study.

Hyun Jun Kong1

  • 1Department of Prosthodontics, College of Dentistry, Wonkwang University, Iksan, Korea.

Journal of Yeungnam Medical Science
|July 26, 2023
PubMed
Summary
This summary is machine-generated.

This study tested if a cloud-based artificial intelligence system could accurately identify different types of dental implants from standard X-ray images. The researchers trained a computer model using thousands of radiographs and found it could reliably distinguish between four specific implant brands. This technology could help dentists quickly identify unknown implants during patient follow-up visits.

Keywords:
Artificial intelligenceCloud computingComputer neural networksDeep learningDental implantsartificial intelligencedental diagnosticsimage classificationcloud computing

Frequently Asked Questions

More Related Videos

Digital Hybrid Model Preparation for Virtual Planning of Reconstructive Dentoalveolar Surgical Procedures
09:10

Digital Hybrid Model Preparation for Virtual Planning of Reconstructive Dentoalveolar Surgical Procedures

Published on: August 5, 2021

1.8K
Dynamic Navigation in Endodontics: Guided Access Cavity Preparation by Means of a Miniaturized Navigation System
07:03

Dynamic Navigation in Endodontics: Guided Access Cavity Preparation by Means of a Miniaturized Navigation System

Published on: May 5, 2022

4.5K

Related Experiment Videos

Last Updated: Jul 21, 2025

Author Spotlight: Advancing CBCT and Digital Dental Image Integration with AI-Assisted Digitization
05:49

Author Spotlight: Advancing CBCT and Digital Dental Image Integration with AI-Assisted Digitization

Published on: February 23, 2024

887
Digital Hybrid Model Preparation for Virtual Planning of Reconstructive Dentoalveolar Surgical Procedures
09:10

Digital Hybrid Model Preparation for Virtual Planning of Reconstructive Dentoalveolar Surgical Procedures

Published on: August 5, 2021

1.8K
Dynamic Navigation in Endodontics: Guided Access Cavity Preparation by Means of a Miniaturized Navigation System
07:03

Dynamic Navigation in Endodontics: Guided Access Cavity Preparation by Means of a Miniaturized Navigation System

Published on: May 5, 2022

4.5K

Area of Science:

  • Dental implant systems diagnostics within biomedical engineering
  • Advanced computational imaging and deep learning applications

Background:

No prior work had resolved the challenge of identifying specific dental implant brands from standard radiographs using automated computational tools. Clinicians often struggle to determine the exact hardware present in patients who received treatment elsewhere. That uncertainty drove the need for reliable, non-invasive diagnostic methods to assist in routine dental care. Prior research has shown that convolutional neural networks can perform complex image recognition tasks with high precision. This gap motivated the development of cloud-based platforms to streamline the training and deployment of such diagnostic models. Researchers have previously explored various manual and semi-automated techniques to categorize medical hardware. However, these traditional approaches often lack the scalability required for busy clinical environments. This study builds upon existing artificial intelligence frameworks to address the identification of diverse implant systems.

Purpose Of The Study:

The study aimed to evaluate the accuracy and clinical usability of implant system classification using automated machine learning on a cloud platform. Researchers sought to determine if artificial intelligence could reliably identify specific dental hardware from standard X-ray images. This investigation addressed the difficulty clinicians face when identifying unknown implants in patients. The team focused on four specific brands to test the model's discriminatory power. By leveraging cloud-based neural architecture search, the authors intended to simplify the development of diagnostic tools. They wanted to assess whether automated processes could match the performance of manual identification methods. This work was motivated by the need for faster, more objective hardware recognition in dental practice. The researchers established a clear framework to measure the effectiveness of their proposed computational solution.

Main Methods:

The review approach involved collecting 4,800 periapical images from electronic medical records for four distinct hardware categories. Researchers manually isolated regions of interest by cropping each file to 400×800 pixels. These processed files were uploaded to a centralized storage environment for automated analysis. The team utilized a cloud-based platform to execute a neural architecture search for model optimization. Eighty percent of the total data served as the training set, while the remaining portions were split equally for validation and testing. A single-label classification architecture was generated to categorize the hardware based on visual patterns. Performance was quantified using standard statistical metrics including precision, recall, and specificity. This systematic workflow ensured that the computational model could effectively learn to distinguish between the selected implant brands.

Main Results:

Key findings from the literature indicate that the cloud-based model achieved an overall accuracy of 0.981. The system demonstrated a precision of 0.963 and a recall of 0.961 across the tested categories. Researchers reported a specificity value of 0.985 for the classification task. The F1 score for the model was recorded at 0.962. Notably, the Osstem TSIII system was identified with 100% accuracy during the evaluation phase. The confusion matrix revealed that the Osstem USII and 3i Osseotite External systems were the most frequently misclassified. These results confirm that the automated approach functions as a fine-tuned convolutional neural network. The data show that the model maintains high performance levels when processing standardized radiographic inputs.

Conclusions:

The authors propose that cloud-based automated machine learning provides a robust solution for identifying dental hardware from radiographic images. Their findings suggest that this technology achieves high performance metrics across multiple classification criteria. The researchers highlight that the model successfully distinguished between four distinct implant types with high reliability. They note that the Osstem TSIII system demonstrated perfect classification accuracy in their testing set. The study indicates that confusion between specific brands remains a challenge for the current neural architecture. The authors suggest that incorporating higher-quality images will be necessary to refine the model's diagnostic capabilities. They emphasize that expanding the dataset to include a wider variety of systems could improve overall clinical utility. The team concludes that this approach represents a viable path toward integrating advanced image recognition into daily dental practice.

The researchers utilized a neural architecture search technology within the Google Cloud platform. This automated process identified the optimal algorithm for classifying periapical radiographs, achieving an overall accuracy of 0.981 and an F1 score of 0.962.

The study focused on four distinct hardware types: Osstem TSIII, Osstem USII, Biomet 3i Osseotite External, and Dentsply Sirona Xive. Each category was represented by 1,200 labeled periapical radiographs to ensure balanced training data.

The authors state that manual cropping of regions of interest to 400×800 pixels was necessary to standardize the input data. This preprocessing step ensured that the neural network focused on relevant anatomical and hardware features during the training phase.

The researchers used a dataset of 4,800 total images. They allocated 80% for training the model, 10% for validation during the learning process, and 10% for final performance testing to ensure unbiased results.

The model achieved an accuracy of 0.981, precision of 0.963, recall of 0.961, and specificity of 0.985. The Osstem TSIII system reached 100% accuracy, while the Osstem USII and 3i Osseotite External systems were frequently misidentified by the software.

The researchers propose that future iterations require higher-quality images from a broader range of manufacturers. They suggest this expansion is vital to improve the model's clinical usability and reduce confusion between similar-looking hardware components.