Oral RadiologyPub Date : 2025-04-01Epub Date: 2024-12-09DOI: 10.1007/s11282-024-00794-y
Berrin Çelik, Muhammed Emin Baslak, Mehmet Zahid Genç, Mahmut Emin Çelik
{"title":"Automated segmentation of dental restorations using deep learning: exploring data augmentation techniques.","authors":"Berrin Çelik, Muhammed Emin Baslak, Mehmet Zahid Genç, Mahmut Emin Çelik","doi":"10.1007/s11282-024-00794-y","DOIUrl":"10.1007/s11282-024-00794-y","url":null,"abstract":"<p><strong>Objectives: </strong>Deep learning has revolutionized image analysis for dentistry. Automated segmentation of dental radiographs is of great importance towards digital dentistry. The performance of deep learning models heavily relies on the quality and diversity of the training data. Data augmentation is a widely used technique implemented in machine learning and deep learning to artificially increase the size and diversity of a training dataset by applying various transformations to the original data.</p><p><strong>Methods: </strong>This work aims to automatically segment implants, prostheses, and fillings in panoramic images using 9 different deep learning segmentation models. Later, it explores the effect of data augmentation methods on segmentation performance of the models. Eight different data augmentation techniques are examined. Performance is evaluated by well-accepted metrics such as intersection over union (IoU) and Dice coefficient.</p><p><strong>Results: </strong>While averaging the segmentation results for the three classes, IoU varies between 0.62 and 0.82 while Dice score is between 0.75 and 0.9 among deep learning models used. Augmentation techniques provided performance improvements of up to 3.37%, 5.75% and 8.75% for implant, prosthesis and filling classes, respectively.</p><p><strong>Conclusions: </strong>Findings reveal that choosing optimal augmentation strategies depends on both model architecture and dental structure type.</p>","PeriodicalId":56103,"journal":{"name":"Oral Radiology","volume":" ","pages":"207-215"},"PeriodicalIF":1.6,"publicationDate":"2025-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142803425","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"The effectiveness of fractal analysis in diagnosing temporomandibular joint disorders: a systematic review of clinical studies.","authors":"Sanjana Santhosh Kumar, Ravleen Nagi, Rachel Chacko, Junad Khan","doi":"10.1007/s11282-024-00791-1","DOIUrl":"10.1007/s11282-024-00791-1","url":null,"abstract":"<p><strong>Objective: </strong>This study aims to evaluate the application and effectiveness of fractal analysis (FA) in assessing temporomandibular joint disorders (TMDs) with dental imaging.</p><p><strong>Methods: </strong>This systematic review, conducted in adherence to PRISMA-P and Cochrane Handbook guidelines, involved a comprehensive search of five electronic indexed databases up to September 15, 2024. The thorough search aimed to ensure the inclusion of all relevant studies on dental imaging with fractal dimension (FD) analysis for TMDs. The risk of bias was performed using the revised QUADAS 2 tool.</p><p><strong>Results: </strong>Out of 342 studies retrieved, 15 met the inclusion criteria and were included in the systematic review. These studies comprised 7 retrospective and 8 prospective nonrandomized clinical studies. Various imaging modalities were used including panoramic, CT, CBCT, and MRI. Most studies reported significantly lower FD values in TMD patients than in controls suggesting FD analysis' potential for detecting early TMJ degenerative changes. However, a few studies did not find significant differences or lacked control groups, highlighting the variability in findings across the research. The overall risk of bias was high regarding the applicability of all included studies.</p><p><strong>Conclusion: </strong>The fractal dimension (FD) analysis of dental images shows potential as a valuable tool for detecting early degenerative changes in temporomandibular disorders (TMDs). It could enhance diagnostic efficiency by providing additional insights from routine radiographs. However, the variability in findings and methodologies underscores the need for further research to validate and standardize these techniques.</p>","PeriodicalId":56103,"journal":{"name":"Oral Radiology","volume":" ","pages":"153-168"},"PeriodicalIF":1.6,"publicationDate":"2025-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142803442","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Oral RadiologyPub Date : 2025-03-29DOI: 10.1007/s11282-025-00819-0
Yuri Nejaim, Gabriela Sofia Noé Bregolin, Natália Hitomi Suekane, Estéfany Figueiredo Gonzalez, Fernando Roberto Ferreira Valente, Rocharles Cavalcante Fontenele
{"title":"Does facial sunscreen usage impact radiographic image quality and radiation dose? An in vitro study.","authors":"Yuri Nejaim, Gabriela Sofia Noé Bregolin, Natália Hitomi Suekane, Estéfany Figueiredo Gonzalez, Fernando Roberto Ferreira Valente, Rocharles Cavalcante Fontenele","doi":"10.1007/s11282-025-00819-0","DOIUrl":"https://doi.org/10.1007/s11282-025-00819-0","url":null,"abstract":"<p><strong>Objective: </strong>To assess whether the use of different types of facial sunscreen influences the quality of radiographic images and the absorbed radiation dose during radiographic acquisitions.</p><p><strong>Methods: </strong>In this in vitro study, two types of facial sunscreens (Bioderma<sup>®</sup>), both with a sun protection factor of 50, were tested: one organic and one inorganic. A polystyrene plate was used, containing a thermoluminescent dosimeter and a photostimulable phosphor plate for the radiographs. The sunscreen was applied to the plate, and five radiographs were taken for each group: control (without sunscreen), organic sunscreen, and inorganic sunscreen. Image quality was assessed by noise, brightness, and uniformity, and the radiation dose was measured in milligrays. The results were compared using one-way analysis of variance with Tukey post-hoc test (α = 5%).</p><p><strong>Results: </strong>Inorganic sunscreen produced images with higher brightness and lower uniformity, with no significant differences in noise. Additionally, this group showed a lower radiation dose (0.50 mGy) compared to the control group (0.60 mGy) and the organic sunscreen (0.58 mGy) (p < 0.05).</p><p><strong>Conclusion: </strong>The inorganic sunscreen altered image quality by increasing brightness and decreasing uniformity, while also reducing the absorbed radiation dose.</p>","PeriodicalId":56103,"journal":{"name":"Oral Radiology","volume":" ","pages":""},"PeriodicalIF":1.6,"publicationDate":"2025-03-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143744546","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Patient knowledge and misconceptions about imaging exams and X-rays: the need for clearer guidance in dental care.","authors":"Geraldo Fagundes Serpa, Lucas Machado Maracci, Gabriela Barbieri Ortigara, Rubens Spin-Neto, Gabriela Salatino Liedke","doi":"10.1007/s11282-025-00817-2","DOIUrl":"https://doi.org/10.1007/s11282-025-00817-2","url":null,"abstract":"<p><strong>Objective: </strong>To assess patients' knowledge and possible misconceptions regarding imaging exams and their association to ionizing radiation.</p><p><strong>Methods: </strong>Patients who sought dental care at the Federal University of Santa Maria Dental School (Brazil) were invited to enroll in the study. Patients filled out a questionnaire that covered demographic information, knowledge regarding imaging exams, X-rays, dental radiographs, and information sources (internet or dentist/physician). Data analysis was performed using descriptive statistics and chi-square test.</p><p><strong>Results: </strong>Two hundred thirty-five participants were enrolled (mean age 44 ± 15 years), of whom 60% were female and 68% had at least 8 years of formal education. Most participants (74.5%) reported knowing what X-rays are. When questioned if the following exams used X-rays, the majority said radiographs used X-rays (91.5%), but mistakes were revealed for tomography (51.7%), mammography (59.4%), and magnetic resonance imaging (MRI) (40.2%). Gender, educational level, and reported knowledge about X-rays were not associated with correct answers (P > 0.05). Younger patients answered more accurately that ultrasound (P = 0.009) and MRI (P = 0.025) do not use X-rays, and older patients correctly associated mammography with X-rays (P < 0.001). Patients whose information source was the internet tended to incorrectly state that mammography (P = 0.007) and tomography (P = 0.063) did not use X-rays.</p><p><strong>Conclusion: </strong>Patients often fail to differentiate the imaging exams that use X-rays, despite reporting that they understand what X-rays are. Dentists should be aware of the misconceptions patients may encounter when using the internet as a source of information and provide clearer guidance about acquisition and purpose of imaging exams, with emphasis on radiation protection.</p>","PeriodicalId":56103,"journal":{"name":"Oral Radiology","volume":" ","pages":""},"PeriodicalIF":1.6,"publicationDate":"2025-03-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143694597","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Oral RadiologyPub Date : 2025-03-17DOI: 10.1007/s11282-025-00815-4
Faezeh Yousefi, Younes Mohammadi, Kimia Nikikhah, Forough Abbasiyan
{"title":"Investigating the effectiveness of MAR algorithm on magnitude of artifacts in CBCT images: a systematic review.","authors":"Faezeh Yousefi, Younes Mohammadi, Kimia Nikikhah, Forough Abbasiyan","doi":"10.1007/s11282-025-00815-4","DOIUrl":"https://doi.org/10.1007/s11282-025-00815-4","url":null,"abstract":"<p><strong>Background: </strong>There has been an increasing interest in the use of implants to treat edentulous patients. In this regard, the use of cone beam computed tomography (CBCT) offers a variety of advantages compared with other imaging methods. However, the creation of beam-hardening artifacts adversely affects the quality of images. To our knowledge, little is known about the actual effectiveness of the Metal Artifact Reduction (MAR) algorithm on image quality improvement.</p><p><strong>Objectives: </strong>The objective of this study is to conduct a systematic review to investigate the effectiveness of the MAR algorithm on the magnitude of artifacts generated in CBCT images.</p><p><strong>Materials and methods: </strong>An electronic search was performed in electronic databases, including PubMed, Scopus, Web of Science, and Google Scholar. For each database, the search strategy was defined specifically. Studies that had the inclusion criteria for this review were imported into Endnote version 20. The risk of bias in the studies included in this systematic review was assessed by two independent reviewers based on the Joanna Briggs Institute (JBI)'s Critical Appraisal checklist. The selected final articles were scored based on the specified checklist. After reviewing selected articles, it was not possible to perform a meta-analysis due to the heterogeneity and multiplicity of the variables, and the studies were included in the systematic review.</p><p><strong>Results: </strong>A total of 4738 studies were identified. After eliminating duplicate and unrelated articles, 10 articles met the inclusion criteria. Results showed that the use of the MAR algorithm in the preparation of CBCT scans reduces the standard deviation (SD) of gray values. However, no definite result was achieved in relation to the contrast-to-noise ratio (CNR). In fact, it cannot be definitively concluded whether the use of the MAR algorithm will increase the CNR.</p><p><strong>Conclusion: </strong>The results of this systematic review demonstrated that we cannot provide a definite answer regarding the effect of the MAR algorithm on reducing the artifacts around dental implants. The explanation is that this factor is affected by many variables, whose change can have a significant effect on the magnitude of artifacts generated in the image.</p>","PeriodicalId":56103,"journal":{"name":"Oral Radiology","volume":" ","pages":""},"PeriodicalIF":1.6,"publicationDate":"2025-03-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143652064","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Image findings and pathological findings of lymphoepithelial cysts with hyper internal echo-a comparison with dermoid cysts.","authors":"Misa Yasaka, Kazutoshi Okamura, Mayumi Shimizu, Koji Sagiyama, Tamotsu Kiyoshima, Shintaro Kawano, Masafumi Moriyama, Toru Chikui","doi":"10.1007/s11282-025-00818-1","DOIUrl":"https://doi.org/10.1007/s11282-025-00818-1","url":null,"abstract":"<p><strong>Objective: </strong>To summarize the ultrasonography (US) and magnetic resonance imaging (MRI) findings of lymphoepithelial cysts (LECs) and dermoid cysts (DCs) and to discuss the hyperinternal echo in LECs based on histopathological findings.</p><p><strong>Methods: </strong>Patients who underwent US and MRI before resection and whose lesions were pathologically diagnosed as LECs or DCs were included (January 2010 to December 2023). Internal echo, convective flow, and Doppler signals were evaluated on US, while on MRI, we measured the apparent diffusion coefficient (ADC).</p><p><strong>Results: </strong>Surprisingly, all lesions (4 LECs and 7 DCs) had hyperinternal echoes. Convective flow and Doppler signals were observed in 2 of the 4 cases of LEC, and the ADC was high in these cases (2.4 × 10<sup>-3</sup> and 2.0 × 10<sup>-3</sup> mm<sup>2</sup>/s). In the two cases where they were not, the ADC was low (1.4 × 10<sup>-3</sup> and 0.81 × 10<sup>-3</sup> mm<sup>2</sup>/s). Various cells, such as lymphocyte cells, foam cells, erythroid cells, and keratin, have been found in the cystic cavity, which are thought to be the cause of hyperecho. Considering the ADC, the presence or absence of convection and Doppler signals is thought to reflect the ease of movement of the contents. In the DCs, neither internal convection nor obvious Doppler signals were observed. ADCs were quite low for cysts, which was a characteristic finding (1.1 ± 0.13<sup>-3</sup> mm<sup>2</sup>/s).</p><p><strong>Conclusion: </strong>It is well known that DCs appear hyperechoic on US owing to the presence of keratin; however, it is necessary to recognize that some LECs are depicted as hyperechoic masses and have convective flow and Doppler signals depending on the content.</p>","PeriodicalId":56103,"journal":{"name":"Oral Radiology","volume":" ","pages":""},"PeriodicalIF":1.6,"publicationDate":"2025-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143634972","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Evaluation of the effects of bisphosphonate therapy on the temporomandibular joint using cone beam computed tomography.","authors":"Bilay Stevanovic Sancar, Rümeysa Şendişçi Gök, Selmi Tunç","doi":"10.1007/s11282-025-00816-3","DOIUrl":"https://doi.org/10.1007/s11282-025-00816-3","url":null,"abstract":"<p><strong>Objective: </strong>The aim of this study was to evaluate the mandibular condyle and articular eminences of patients diagnosed with medication-related osteonecrosis of the jawbone (MRONJ) and those who use bisphosphonates (BP) but do not have MRONJ findings, by comparing them with a healthy control group.</p><p><strong>Methods: </strong>The cone beam computed tomography (CBCT) images of 20 patients (15 male, 5 female) who were diagnosed with MRONJ and 20 patients (16 male and 4 female) who were using bisphosphonates but had no MRONJ were included in the study. A control group consisted of 20 age- and gender-matched patients (13 male, 7 female) who had no clinical complaints or signs and symptoms of TMD or rheumatic disease. The imaging characteristics relating to osseous changes of the condyles and in the articular fossa, such as flattening, erosion, osteophytes, subcortical sclerosis, subchondral cyst, loose joint bodies, joint space narrowing, flattening of the articular eminence, and resorption in the articular eminence, were evaluated. SPSS Statistics 20 software was used to statistically evaluate the data collected within the scope of the investigated variables. The results of the analyses were evaluated at 0.05.</p><p><strong>Results: </strong>Analysis of temporomandibular joint osseous changes in the patient and control groups found statistically significantly higher subcortical sclerosis in both the patient group diagnosed with MRONJ and the patient group using BP, compared with the control group. In addition, a statistically significant difference was found between the patient groups and the control group in terms of condylar flattening, articular eminence flattening, and resorption.</p><p><strong>Conclusion: </strong>Understanding the changes in physiological bone remodeling of the mandibular condyle after BP administration is crucial in the development of future treatment modalities for degenerative TMJ diseases that cause condylar bone resorption.</p>","PeriodicalId":56103,"journal":{"name":"Oral Radiology","volume":" ","pages":""},"PeriodicalIF":1.6,"publicationDate":"2025-03-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143617931","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Association between CT image findings and hypoesthesia in the mental region and lower lip after dentigerous cyst removal: a retrospective study and literature review.","authors":"Shinichiro Yamada, Hizuru Osanai, Kozue Mori, Noriyuki Sugino, Akira Taguchi","doi":"10.1007/s11282-025-00813-6","DOIUrl":"https://doi.org/10.1007/s11282-025-00813-6","url":null,"abstract":"<p><strong>Objectives: </strong>To evaluate the role of computed tomography (CT)/cone-beam computed beam CT (CBCT) in association with hypoesthesia of the mental region and lower lip during the surgical removal of DCs.</p><p><strong>Methods: </strong>A retrospective cohort study was performed involving 39 impacted mandibular DC cases between February 2011 and March 2022. We studied the association of hypoesthesia in relation to the age and sex of the patient and to growth direction and expansion of the DC on the CT/CBCT image findings. The preoperative CT/CBCT image findings were statistically analyzed in accordance with diagnoses of DCs confirmed based on pathology. Fisher's exact test was used to test the relationship between hypoesthesia and CT/CBCT imaging findings. In addition, PubMed and PubMed Central were searched using keywords such as \"hypoesthesia\" to find relevant studies. Reports of hypoesthesia of the mental region and lower lip were derived from the selected literature.</p><p><strong>Results: </strong>The frequency of hypoesthesia recorded was 35.9%. No statistically significant association was observed for all hypoesthesia and CT/CBCT image findings.</p><p><strong>Conclusions: </strong>Imaging evaluation was performed in 39 dentigerous cysts, and a literature review was performed. We found that CT/CBCT alone does not provide sufficiently reliable images required for predicting nerve lesions. Future prospective studies are required.</p>","PeriodicalId":56103,"journal":{"name":"Oral Radiology","volume":" ","pages":""},"PeriodicalIF":1.6,"publicationDate":"2025-03-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143558867","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Evaluation of the effect of reducing metal artifacts in multi-detector CT imaging of zirconia and titanium implants.","authors":"Hideki Suito, Hozumi Yoshihara, Naoki Maeda, Ryosuke Kasai, Tadashi Inoue, Masafumi Amano, Yuuri Oku, Keiko Fujimoto, Kan Nagao","doi":"10.1007/s11282-025-00814-5","DOIUrl":"https://doi.org/10.1007/s11282-025-00814-5","url":null,"abstract":"<p><strong>Objectives: </strong>Zirconia oral implants produce larger artifacts on computed tomography (CT) compared to titanium oral implants. This study aimed to compare the effectiveness of metal artifact reduction (MAR) software and dual-energy (DE) methods, which are used in multi-detector CT (MDCT) to reduce the artifacts produced by titanium and zirconia oral implants.</p><p><strong>Methods: </strong>Oral implants were placed in the lower right second premolar and second molar regions of an acrylic mandible model and evaluated using MDCT. For MAR, Single Energy Metal Artifact Reaction (SEMAR) images were obtained, and for DE, images were obtained at 135 keV. The evaluation was based on 21 regions of interest (ROIs) with sides of 2.8 mm, set at 15, 25, and 35 mm from the center (the lower-right first premolar), with 65°, 90°, 115°, and 145° angles set on the buccal and lingual sides. The evaluation was based on an Artifact Index (AI).</p><p><strong>Results: </strong>The lingual aspect of the AI tended to be larger compared to the buccal aspect, regardless of the image-processing method. In addition, the SEMAR method tended to have a greater artifact reduction effect than the DE method. The distortion of the mandibular model and occurrence of new artifacts were confirmed using SEMAR.</p><p><strong>Conclusions: </strong>SEMAR had higher artifact reduction effects on metal artifacts than the DE method. Although SEMAR and dual-energy methods can effectively reduce metal artifacts, they also cause false images and areas that cannot be observed. Therefore, diagnostic and treatment planning, including imaging, is essential before artifact reduction processing.</p>","PeriodicalId":56103,"journal":{"name":"Oral Radiology","volume":" ","pages":""},"PeriodicalIF":1.6,"publicationDate":"2025-03-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143558902","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Oral RadiologyPub Date : 2025-02-28DOI: 10.1007/s11282-025-00812-7
Hatice Ahsen Deniz, İbrahim Şevki Bayrakdar, Rana Nalçacı, Kaan Orhan
{"title":"Segmentation of the nasopalatine canal and detection of canal furcation status with artificial intelligence on cone-beam computed tomography images.","authors":"Hatice Ahsen Deniz, İbrahim Şevki Bayrakdar, Rana Nalçacı, Kaan Orhan","doi":"10.1007/s11282-025-00812-7","DOIUrl":"https://doi.org/10.1007/s11282-025-00812-7","url":null,"abstract":"<p><strong>Objectives: </strong>The nasopalatine canal (NPC) is an anatomical formation with varying morphology. NPC can be visualized using the cone-beam computed tomography (CBCT). Also, CBCT has been used in many studies on artificial intelligence (AI). The \"You only look once\" (YOLO) is an AI framework that stands out with its speed. This study compared the observer and AI regarding the NPC segmentation and assessment of the NPC furcation status in CBCT images.</p><p><strong>Methods: </strong>In this study, axial sections of 200 CBCT images were used. These images were labeled and evaluated for the absence or presence of the NPC furcation. These images were then divided into three; 160 images were used as the training dataset, 20 as the validation dataset, and 20 as the test dataset. The training was performed by making 800 epochs using the YOLOv5x-seg model.</p><p><strong>Results: </strong>Sensitivity, Precision, F1 score, IoU, mAP, and AUC values were determined for NPC detection, segmentation, and classification of the YOLOv5x-seg model. The values were found to be 0.9680, 0.9953, 0.9815, 0.9636, 0.7930, and 0.8841, respectively, for the group with the absence of the NPC furcation; and 0.9827, 0.9975, 0.9900, 0.9803, 0.9637, and 0.9510, for the group with the presence of the NPC furcation.</p><p><strong>Conclusions: </strong>Our results showed that even when the YOLOv5x-seg model is trained with the NPC furcation and fewer datasets, it achieves sufficient prediction accuracy. The segmentation feature of the YOLOv5 algorithm, which is based on an object detection algorithm, has achieved quite successful results despite its recent development.</p>","PeriodicalId":56103,"journal":{"name":"Oral Radiology","volume":" ","pages":""},"PeriodicalIF":1.6,"publicationDate":"2025-02-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143532254","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}