Welcome to PracticeUpdate! We hope you are enjoying access to a selection of our top-read and most recent articles. Please register today for a free account and gain full access to all of our expert-selected content.
Already Have An Account? Log in Now
The Convolutional Neural Network Is Effective in Identifying Fused-Rooted Mandibular Second Molars on Radiographs
TAKE-HOME MESSAGE
- Making a 3D diagnosis using 2D radiographs is sometimes necessary but challenging, and this study explored the possibility of using a convolutional neural network (CNN), a deep learning approach, to diagnose fused-rooted mandibular second molars (MSMs) on 2D radiographs. Micro-CT reconstruction images of extracted fused-rooted MSMs were used as ground truth. The CNN was applied to radiographic images to detect MSMs, and the detection of MSMs by endodontic residents was the comparison group. The CNN resulted in an overall detection accuracy of 79.25% compared with a detection accuracy of 60.38% by endodontic residents.
- The CNN was effective in identifying fused-rooted MSMs on radiographs and, therefore, could significantly aid the clinician in making 3D diagnoses using 2D radiographs.
abstract
This abstract is available on the publisher's site.
Access this abstract now Full Text Available for ClinicalKey SubscribersINTRODUCTION
Understanding the intricate anatomical morphology of fused-rooted mandibular second molars (MSMs) is essential for root canal treatment. The present study utilized a deep learning approach to identify the three-dimensional root canal morphology of MSMs from two-dimensional X-ray images.
METHODS
A total of 271 fused-rooted MSMs were included in the study. Micro-computed tomography reconstruction images and two-dimensional X-ray projection images were obtained. The ground truth of three-dimensional root canal morphology was determined through micro-computed tomography images, which were classified into merging, symmetrical, and asymmetrical types. To amplify the X-ray image dataset, traditional augmentation techniques from the python package Augmentor and a multiangle projection method were employed. Identification of root canal morphology was conducted using the pretrained VGG19, ResNet18, ResNet50, and EfficientNet-b5 on X-ray images. The classification results from convolutional neural networks (CNNs) were then compared with those performed by endodontic residents.
RESULTS
The multiangle projection augmentation method outperformed the traditional approach in all CNNs except for EfficientNet-b5. ResNet18 combined with the multiangle projection method outperformed all other combinations, with an overall accuracy of 79.25%. In specific classifications, accuracies of 81.13%, 86.79%, and 90.57% were achieved for merging, symmetrical, and asymmetrical types, respectively. Notably, CNNs surpassed endodontic residents in classification performance; the average accuracy for endodontic residents was only 60.38% (P < .05).
CONCLUSIONS
CNNs were more effective than endodontic residents in identifying the three-dimensional root canal morphology of MSMs. The result indicates that CNNs possess the capacity to employ two-dimensional images effectively in aiding three-dimensional diagnoses.
Additional Info
Disclosure statements are available on the authors' profiles:
Identification of Root Canal Morphology in Fused-rooted Mandibular Second Molars From X-ray Images Based on Deep Learning
J Endod 2024 May 29;[EPub Ahead of Print], W Wu, S Chen, P Chen, M Chen, Y Yang, Y Gao, J Hu, J MaFrom MEDLINE®/PubMed®, a database of the U.S. National Library of Medicine.
Endodontic therapy is often challenging owing to the complex root canal morphology and internal intricacies, as revealed by micro-CT (mCT) studies. Variations in internal anatomies, such as C-shaped canals, second mesiobuccal canals, isthmuses, and apical canal merging and separations, have been associated with the race, sex, and age of the patient. More recently, several technologies related to convolutional neural networks (CNNs), which can extrapolate data from digital dental radiographs, have shown some promising results in terms of the accuracy of diagnoses made using 2D radiographs.
A recent study by Wu et al in 2024 aimed at utilizing a deep learning framework to identify the 3D root canal morphology of mandibular second molars (MSMs) on 2D radiographic images, using morphological data from mCT images as ground truth. This study looked at a Chinese population involving a set of 271 fused-rooted MSMs, with 109 merging-type, 119 symmetrical-type, and 43 asymmetrical-type teeth and variations in C-shaped canals. The mCT images were acquired and computed with MeVisLab software. Additionally, radiographic evaluations were performed by endodontic residents, which were later compared with those obtained from CNN models. The results of this study showed that ResNet18 had higher overall accuracy in diagnostic abilities than residents, proving that the CNN demonstrates a significant improvement in classification accuracy. This model was able to identify apical separation, asymmetrical cervical isthmus, apical isthmus, and apical merge, all of which are related to 87.48% of the model’s decision-making process.
In conclusion, the CNN's ability to identify the 3D root canal morphology of fused-rooted MSMs from 2D radiographs is impressive compared with the visual capabilities of humans. Hence, clinicians should embrace these CNN-based technologies, as they show promising results as an adjunct diagnostic aid. In addition, these technologies also have great potential in education programs because they heighten the diagnostic value of radiographs, providing substantial benefits to both patients and the new generation of dentists.