Application of deep learning and feature selection technique on external root resorption identification on CBCT images
Abstract Background Artificial intelligence has been proven to improve the identification of various maxillofacial lesions. The aim of the current study is two-fold: to assess the performance of four deep learning models (DLM) in external root resorption (ERR) identification and to assess the effect...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
BMC
2024-02-01
|
| Series: | BMC Oral Health |
| Subjects: | |
| Online Access: | https://doi.org/10.1186/s12903-024-03910-w |
| _version_ | 1797272959268683776 |
|---|---|
| author | Nor Hidayah Reduwan Azwatee Abdul Abdul Aziz Roziana Mohd Razi Erma Rahayu Mohd Faizal Abdullah Seyed Matin Mazloom Nezhad Meghna Gohain Norliza Ibrahim |
| author_facet | Nor Hidayah Reduwan Azwatee Abdul Abdul Aziz Roziana Mohd Razi Erma Rahayu Mohd Faizal Abdullah Seyed Matin Mazloom Nezhad Meghna Gohain Norliza Ibrahim |
| author_sort | Nor Hidayah Reduwan |
| collection | DOAJ |
| description | Abstract Background Artificial intelligence has been proven to improve the identification of various maxillofacial lesions. The aim of the current study is two-fold: to assess the performance of four deep learning models (DLM) in external root resorption (ERR) identification and to assess the effect of combining feature selection technique (FST) with DLM on their ability in ERR identification. Methods External root resorption was simulated on 88 extracted premolar teeth using tungsten bur in different depths (0.5 mm, 1 mm, and 2 mm). All teeth were scanned using a Cone beam CT (Carestream Dental, Atlanta, GA). Afterward, a training (70%), validation (10%), and test (20%) dataset were established. The performance of four DLMs including Random Forest (RF) + Visual Geometry Group 16 (VGG), RF + EfficienNetB4 (EFNET), Support Vector Machine (SVM) + VGG, and SVM + EFNET) and four hybrid models (DLM + FST: (i) FS + RF + VGG, (ii) FS + RF + EFNET, (iii) FS + SVM + VGG and (iv) FS + SVM + EFNET) was compared. Five performance parameters were assessed: classification accuracy, F1-score, precision, specificity, and error rate. FST algorithms (Boruta and Recursive Feature Selection) were combined with the DLMs to assess their performance. Results RF + VGG exhibited the highest performance in identifying ERR, followed by the other tested models. Similarly, FST combined with RF + VGG outperformed other models with classification accuracy, F1-score, precision, and specificity of 81.9%, weighted accuracy of 83%, and area under the curve (AUC) of 96%. Kruskal Wallis test revealed a significant difference (p = 0.008) in the prediction accuracy among the eight DLMs. Conclusion In general, all DLMs have similar performance on ERR identification. However, the performance can be improved by combining FST with DLMs. |
| first_indexed | 2024-03-07T14:36:48Z |
| format | Article |
| id | doaj.art-5b3d5241de7540509a35f20a7a4ab99d |
| institution | Directory Open Access Journal |
| issn | 1472-6831 |
| language | English |
| last_indexed | 2024-03-07T14:36:48Z |
| publishDate | 2024-02-01 |
| publisher | BMC |
| record_format | Article |
| series | BMC Oral Health |
| spelling | doaj.art-5b3d5241de7540509a35f20a7a4ab99d2024-03-05T20:34:35ZengBMCBMC Oral Health1472-68312024-02-0124111010.1186/s12903-024-03910-wApplication of deep learning and feature selection technique on external root resorption identification on CBCT imagesNor Hidayah Reduwan0Azwatee Abdul Abdul Aziz1Roziana Mohd Razi2Erma Rahayu Mohd Faizal Abdullah3Seyed Matin Mazloom Nezhad4Meghna Gohain5Norliza Ibrahim6Department of Oral and Maxillofacial Clinical Sciences, Faculty of Dentistry, Universiti MalayaDepartment of Restorative Dentistry, Faculty of Dentistry, Universiti MalayaDepartment of Pediatric Dentistry and Orthodontic, Faculty of Dentistry, Universiti MalayaDepartment of Artificial Intelligence, Faculty of Computer Science and Information Technology, Universiti MalayaDepartment of Artificial Intelligence, Faculty of Computer Science and Information Technology, Universiti MalayaDepartment of Oral and Maxillofacial Clinical Sciences, Faculty of Dentistry, Universiti MalayaDepartment of Oral and Maxillofacial Clinical Sciences, Faculty of Dentistry, Universiti MalayaAbstract Background Artificial intelligence has been proven to improve the identification of various maxillofacial lesions. The aim of the current study is two-fold: to assess the performance of four deep learning models (DLM) in external root resorption (ERR) identification and to assess the effect of combining feature selection technique (FST) with DLM on their ability in ERR identification. Methods External root resorption was simulated on 88 extracted premolar teeth using tungsten bur in different depths (0.5 mm, 1 mm, and 2 mm). All teeth were scanned using a Cone beam CT (Carestream Dental, Atlanta, GA). Afterward, a training (70%), validation (10%), and test (20%) dataset were established. The performance of four DLMs including Random Forest (RF) + Visual Geometry Group 16 (VGG), RF + EfficienNetB4 (EFNET), Support Vector Machine (SVM) + VGG, and SVM + EFNET) and four hybrid models (DLM + FST: (i) FS + RF + VGG, (ii) FS + RF + EFNET, (iii) FS + SVM + VGG and (iv) FS + SVM + EFNET) was compared. Five performance parameters were assessed: classification accuracy, F1-score, precision, specificity, and error rate. FST algorithms (Boruta and Recursive Feature Selection) were combined with the DLMs to assess their performance. Results RF + VGG exhibited the highest performance in identifying ERR, followed by the other tested models. Similarly, FST combined with RF + VGG outperformed other models with classification accuracy, F1-score, precision, and specificity of 81.9%, weighted accuracy of 83%, and area under the curve (AUC) of 96%. Kruskal Wallis test revealed a significant difference (p = 0.008) in the prediction accuracy among the eight DLMs. Conclusion In general, all DLMs have similar performance on ERR identification. However, the performance can be improved by combining FST with DLMs.https://doi.org/10.1186/s12903-024-03910-wExternal root resorptionCone beam computed tomographyArtificial intelligenceDeep learningFeature selection techniqueClassification |
| spellingShingle | Nor Hidayah Reduwan Azwatee Abdul Abdul Aziz Roziana Mohd Razi Erma Rahayu Mohd Faizal Abdullah Seyed Matin Mazloom Nezhad Meghna Gohain Norliza Ibrahim Application of deep learning and feature selection technique on external root resorption identification on CBCT images BMC Oral Health External root resorption Cone beam computed tomography Artificial intelligence Deep learning Feature selection technique Classification |
| title | Application of deep learning and feature selection technique on external root resorption identification on CBCT images |
| title_full | Application of deep learning and feature selection technique on external root resorption identification on CBCT images |
| title_fullStr | Application of deep learning and feature selection technique on external root resorption identification on CBCT images |
| title_full_unstemmed | Application of deep learning and feature selection technique on external root resorption identification on CBCT images |
| title_short | Application of deep learning and feature selection technique on external root resorption identification on CBCT images |
| title_sort | application of deep learning and feature selection technique on external root resorption identification on cbct images |
| topic | External root resorption Cone beam computed tomography Artificial intelligence Deep learning Feature selection technique Classification |
| url | https://doi.org/10.1186/s12903-024-03910-w |
| work_keys_str_mv | AT norhidayahreduwan applicationofdeeplearningandfeatureselectiontechniqueonexternalrootresorptionidentificationoncbctimages AT azwateeabdulabdulaziz applicationofdeeplearningandfeatureselectiontechniqueonexternalrootresorptionidentificationoncbctimages AT rozianamohdrazi applicationofdeeplearningandfeatureselectiontechniqueonexternalrootresorptionidentificationoncbctimages AT ermarahayumohdfaizalabdullah applicationofdeeplearningandfeatureselectiontechniqueonexternalrootresorptionidentificationoncbctimages AT seyedmatinmazloomnezhad applicationofdeeplearningandfeatureselectiontechniqueonexternalrootresorptionidentificationoncbctimages AT meghnagohain applicationofdeeplearningandfeatureselectiontechniqueonexternalrootresorptionidentificationoncbctimages AT norlizaibrahim applicationofdeeplearningandfeatureselectiontechniqueonexternalrootresorptionidentificationoncbctimages |