Microbial Dental Plaque Analysis in Young Permanent Teeth Using Deep Learning

NCT06603233 | Completed

• 1 other identifier: AnkaraMedipolU-DNT-BCT-01
• Study Type: interventional
• Target: 31
• Locations: 1 country
• Sites: 1

Brief Summary

Background: Dental plaque contributes to a number of common oral conditions such as caries, gingivitis, and periodontitis. As a result, detection and management of plaque is of great importance for the oral health of individuals. The primary objectives of this study were to design a deep learning model for the detection and segmentation of plaque in young permanent teeth and to evaluate the diagnostic accuracy of the model. Methods: The dataset contains 506 dental images from 31 patients aged 8 to 13 years. Six state-of-the-art models were trained and tested using this dataset. The U-Net Transformer model was compared with three dentists for clinical applicability using 35 randomly selected images from the test set.

Conditions

Dental Plaque (Diagnosis); Deep Learning

Keywords

Dental Plaque; Deep Learning; AI Model; Pediatric Dentistry

Outcome Measures

Primary Outcomes (2)

• Determination of IoU and Dice Coefficient values among six state-of-the-AI models

  The IoU score, which computes the ratio between the intersection and the union of two sets, is commonly used to evaluate the accuracy of prediction on semantic segmentation. DeepLabV3+, Mask R-CNN (Detectron2), YOLOv8, U-Net, Super Vision U-net and U-Net Transformer were trained on 354 images and tested on 79 images. IoU and Dice Coefficient values were established among six state-of-the-AI models. As the IoU score increases, the prediction score increases. As the score increases, it becomes more distinctive in determining the model that gives results closest to the correct result.

  two weeks

• Prediction scores of the dentists and U-Net Transformer on 35 test images

  The prediction scores of the three dentists and the AI model (U-Net Transformer) on 35 test images

  two weeks

Secondary Outcomes (1)

• T-test results comparing the AI model and the three dentists

  two weeks

Study Arms (2)

Deep Learning Models Group

EXPERIMENTAL

As artificial intelligence models, DeepLabV3+, Mask R-CNN (Detectron2), YOLOv8, U-Net, Super Vision U-net and U-Net Transformer models, which are state-of-the-art in semantic segmentation, were selected.

Diagnostic Test: Deep Learning Models

The Difference Between The AI Model (U-Net Transformer) and Dentists Group

ACTIVE COMPARATOR

Using the prior knowledge (α = 0.05, β = 0.2) and an effect size of 0.61, the actual power of the comparison between the AI model (U-Net Transformer) and dentists on 34 test images is at least 80%, which is deemed sufficient. Therefore, randomly selected 35 images on the test dataset were labeled by three dentists without seeing the ground truth and were predicted by the AI model. Then, the intersection over union (IoU) score of these labeled and predicted images were calculated. The IoU score, which computes the ratio between the intersection and the union of two sets, is commonly used to evaluate the accuracy of prediction on semantic segmentation. To confirm clinical feasibility, three t-tests, which evaluates the difference between the means of two variables, were applied to IoU scores of dentists and IoU scores of the AI model and a p value \< .05 was considered statistically significant.

Diagnostic Test: The Difference Between The AI Model and Dentists Group

Interventions

The Difference Between The AI Model and Dentists Group DIAGNOSTIC_TEST

The clinical feasibility of the best performing model, statistical hypothesis tests are performed that compares the predictions of the AI model with the assessments from three dentists.

Also known as: The AI Model and Dentists Group

The Difference Between The AI Model (U-Net Transformer) and Dentists Group

Deep Learning Models DIAGNOSTIC_TEST

DeepLabV3+, Mask R-CNN (Detectron2), YOLOv8, U-Net, Super Vision U-net and U-Net Transformer were trained on 354 images and tested on 79 images.

Also known as: The Architecture of Deep Learning Models

Deep Learning Models Group

Eligibility Criteria

• Age: 8 Years - 13 Years
• Sex: all
• Healthy Volunteers: Yes
• Age Groups: Child (0-17)

You may qualify if:

• Anterior young permanent teeth

You may not qualify if:

• Anterior young permanent teeth exhibiting disruptions in enamel tissue integrity such as decay
• Hypoplasia, hypomineralization
• Restored and prosthetically treated teeth
• Young permanent teeth located in the posterior region
• Primary teeth

Sponsors & Collaborators

• Ankara Medipol University: lead

Study Sites (1)

Banu Çiçek Tez

Istanbul, Üsküdar, Turkey (Türkiye)

Location

Related Publications (4)

• Liu L, Xu J, Huan Y, Zou Z, Yeh SC, Zheng LR. A Smart Dental Health-IoT Platform Based on Intelligent Hardware, Deep Learning, and Mobile Terminal. IEEE J Biomed Health Inform. 2020 Mar;24(3):898-906. doi: 10.1109/JBHI.2019.2919916. Epub 2019 Jun 7.

  PMID: 31180873 BACKGROUND

• You W, Hao A, Li S, Wang Y, Xia B. Deep learning-based dental plaque detection on primary teeth: a comparison with clinical assessments. BMC Oral Health. 2020 May 13;20(1):141. doi: 10.1186/s12903-020-01114-6.

  PMID: 32404094 BACKGROUND

• Li S, Guo Y, Pang Z, Song W, Hao A, Xia B, Qin H. Automatic Dental Plaque Segmentation Based on Local-to-Global Features Fused Self-Attention Network. IEEE J Biomed Health Inform. 2022 May;26(5):2240-2251. doi: 10.1109/JBHI.2022.3141773. Epub 2022 May 5.

  PMID: 35015655 BACKGROUND

• Tez BC, Guzel Y, Kiziltan Eliacik BB, Aydin Z. Deep-Learning-Based AI-Model for Predicting Dental Plaque in the Young Permanent Teeth of Children Aged 8-13 Years. Children (Basel). 2025 Apr 7;12(4):475. doi: 10.3390/children12040475.

  PMID: 40310101 DERIVED

Related Links

• reference 11
• Reference 13
• reference 16
• reference 18
• reference 19
• reference 20
• reference 21

MeSH Terms

Conditions

Dental Plaque; Disease

Condition Hierarchy (Ancestors)

Dental Deposits; Tooth Diseases; Stomatognathic Diseases; Pathologic Processes; Pathological Conditions, Signs and Symptoms

Study Officials

• Banu Çiçek Tez, Ph.D

  Ankara Medipol University

  PRINCIPAL INVESTIGATOR

Study Design

• Study Type: interventional
• Phase: not applicable
• Allocation: NON RANDOMIZED
• Masking: TRIPLE
• Who Masked: CARE PROVIDER, INVESTIGATOR, OUTCOMES ASSESSOR
• Purpose: DIAGNOSTIC
• Intervention Model: FACTORIAL
• Sponsor Type: OTHER
• Responsible Party: PRINCIPAL INVESTIGATOR
• PI Title: Assistant Professor

Model Details: Using the prior knowledge (α = 0.05, β = 0.2) and an effect size of 0.61, the actual power of the comparison between the AI model and dentists on 34 test images is at least 80%, which is deemed sufficient. Therefore, randomly selected 35 images on the test dataset were labeled by three dentists without seeing the ground truth and were predicted by the AI model. Then, the intersection over union (IoU) score of these labeled and predicted images were calculated. The IoU score, which computes the ratio between the intersection and the union of two sets, is commonly used to evaluate the accuracy of prediction on semantic segmentation. AI model's dental plaque predictions compared with predictions from three dentists.

Study Record Dates

• First Submitted: June 30, 2024
• First Posted: September 19, 2024
• Study Start: June 1, 2023
• Primary Completion: November 1, 2023
• Study Completion: November 1, 2023
• Last Updated: September 19, 2024

Record last verified: 2024-09

Data Sharing

• IPD Sharing: Will not share

Locations

TU (1)