How well do large language model-based chatbots perform in oral and maxillofacial radiology?

OBJECTIVES: This study evaluated the performance of four large language model (LLM)-based chatbots by comparing their test results with those of dental students on an oral and maxillofacial radiology examination. METHODS: ChatGPT, ChatGPT Plus, Bard, and Bing Chat were tested on 52 questions from regular dental college examinations. These questions were categorized into three educational content areas: basic knowledge, imaging and equipment, and image interpretation. They were also classified as multiple-choice questions (MCQs) and short-answer questions (SAQs). The accuracy rates of the chatbots were compared with the performance of students, and further analysis was conducted based on the educational content and question type. RESULTS: The students' overall accuracy rate was 81.2%, while that of the chatbots varied: 50.0% for ChatGPT, 65.4% for ChatGPT Plus, 50.0% for Bard, and 63.5% for Bing Chat. ChatGPT Plus achieved a higher accuracy rate for basic knowledge than the students (93.8% vs. 78.7%). However, all chatbots performed poorly in image interpretation, with accuracy rates below 35.0%. All chatbots scored less than 60.0% on MCQs, but performed better on SAQs. CONCLUSIONS: The performance of chatbots in oral and maxillofacial radiology was unsatisfactory. Further training using specific, relevant data derived solely from reliable sources is required. Additionally, the validity of these chatbots' responses must be meticulously verified. ADVANCES IN KNOWLEDGE: This study is the first in the field of oral and maxillofacial radiology to assess the knowledge levels of four chatbots. We recommend further training in this domain for all chatbots, given their unsatisfactory performance.

Perceptions and attitudes of dental students and dentists in South Korea toward artificial intelligence: a subgroup analysis based on professional seniority.

BACKGROUND: This study explored dental students' and dentists' perceptions and attitudes toward artificial intelligence (AI) and analyzed differences according to professional seniority. METHODS: In September to November 2022, online surveys using Google Forms were conducted at 2 dental colleges and on 2 dental websites. The questionnaire consisted of general information (8 or 10 items) and participants' perceptions, confidence, predictions, and perceived future prospects regarding AI (17 items). A multivariate logistic regression analysis was performed on 4 questions representing perceptions and attitudes toward AI to identify highly influential factors according to position, age, sex, residence, and self-reported knowledge level about AI of respondents. Participants were reclassified into 2 subgroups based on students' years in school and 4 subgroups based on dentists' years of experience. The chi-square test or Fisher's exact test was used to determine differences between dental students and dentists and between subgroups for all 17 questions. RESULTS: The study included 120 dental students and 96 dentists. Participants with high level of AI knowledge were more likely to be interested in AI compared to those with moderate or low level (adjusted OR 24.345, p < 0.001). Most dental students (60.8%) and dentists (67.7%) predicted that dental AI would complement human limitations. Dental students responded that they would actively use AI in almost all cases (40.8%), while dentists responded that they would use AI only when necessary (44.8%). Dentists with 11-20 years of experience were the most likely to disagree that AI could outperform skilled dentists (50.0%), and respondents with longer careers had higher response rates regarding the need for AI education in schools. CONCLUSIONS: Knowledge level about AI emerged as the factor influencing perceptions and attitudes toward AI, with both dental students and dentists showing similar views on recognizing the potential of AI as an auxiliary tool. However, students' and dentists' willingness to use AI differed. Although dentists differed in their confidence in the abilities of AI, all dentists recognized the need for education on AI. AI adoption is becoming a reality in dentistry, which requires proper awareness, proper use, and comprehensive AI education.

Attention-guided jaw bone lesion diagnosis in panoramic radiography using minimal labeling effort.

Developing a deep-learning-based diagnostic model demands extensive labor for medical image labeling. Attempts to reduce the labor often lead to incomplete or inaccurate labeling, limiting the diagnostic performance of models. This paper (i) constructs an attention-guiding framework that enhances the diagnostic performance of jaw bone pathology by utilizing attention information with partially labeled data; (ii) introduces an additional loss to minimize the discrepancy between network attention and its label; (iii) introduces a trapezoid augmentation method to maximize the utility of minimally labeled data. The dataset includes 716 panoramic radiograph data for jaw bone lesions and normal cases collected and labeled by two radiologists from January 2019 to February 2021. Experiments show that guiding network attention with even 5% of attention-labeled data can enhance the diagnostic accuracy for pathology from 92.41 to 96.57%. Furthermore, ablation studies reveal that the proposed augmentation methods outperform prior preprocessing and augmentation combinations, achieving an accuracy of 99.17%. The results affirm the capability of the proposed framework in fine-grained diagnosis using minimally labeled data, offering a practical solution to the challenges of medical image analysis.

Evaluation of masticatory muscles in temporomandibular joint disorder patients using quantitative MRI fat fraction analysis-Could it be a biomarker?

Temporomandibular joint disorders (TMDs) are closely related to the masticatory muscles, but objective and quantitative methods to evaluate muscle are lacking. IDEAL-IQ, a type of chemical shift-encoded magnetic resonance imaging (CSE-MRI), can quantify the fat fraction (FF). The purpose of this study was to develop an MR IDEAL-IQ-based method for quantitative muscle diagnosis in TMD patients. A total of 65 patients who underwent 3 T MRI scans, including CSE-MRI sequences, were retrospectively included. MRI diagnoses and clinical data were reviewed. There were 19 patients in the normal group and 46 patients in the TMD group with unilateral disc displacement. The TMD group was subdivided into those with and without clenching. The right and left FF values of the masseter, medial, and lateral pterygoid muscles were measured twice by two oral radiologists on CSE-MRI, and the average value was used. FF measurements using CSE-MRI showed excellent intra- and inter-observer agreement (ICC > 0.889 for both). There were no statistically significant differences between the right and left FF values in the masseter, medial pterygoid, and lateral pterygoid of the normal group (p > 0.05). A statistically significant difference was found in the TMD group without clenching, in which the masseter muscle had a statistically significantly lower FF value on the disc displacement side (3.94 ± 1.61) than on the normal side (4.52 ± 2.24) (p < 0.05). CSE-MRI, which can reproducibly quantify muscle FF values, is expected to be a biomarker for objective muscle evaluation in TMD patients. The masseter muscle is expected to be particularly useful compared to other masticatory muscles, but further research is needed.

CT and MR imaging findings of head and neck chondrosarcoma.

OBJECTIVES: This study investigated the imaging features of head and neck chondrosarcoma (HNCS) according to its origin and pathologic subtype. METHODS: Patients who were pathologically diagnosed with HNCS between January 2000 and April 2022 were retrospectively reviewed. Lesions were classified based on their origin and pathologic subtype. The size and margin were evaluated on the image. Internal calcification and the effects on adjacent bone were assessed using computed tomography (CT) images, while signal intensity and contrast enhancement patterns were analyzed using magnetic resonance (MR) imaging. RESULTS: Thirteen HNCSs were included in this study: 8 bone tumors (61.5%) and 5 soft tissue tumors (38.5%). The bone tumors were pathologically diagnosed as conventional (n = 5) and mesenchymal type (n = 3). Soft tissue tumors were defined as myxoid type. The main symptoms were swelling (90.9%) and pain (72.7%). The lesions measured 4.5 cm on average. The margins showed benign and well-defined except for the mesenchymal type. On CT, most bone tumors (75%) showed internal calcification with remodeling or destruction of the adjacent bone. No soft tissue tumors, except one case, showed internal calcification or destruction of the adjacent bone. MR imaging features were non-specific (T2 high signal intensity and contrast enhancement). CONCLUSIONS: HCNS showed various imaging findings according to their origin and pathologic subtype. HNCS should be differentiated if a bone tumor shows internal calcification and affects the adjacent bone. When diagnosing slow-growing soft tissue tumors, even if low possibility, HNCS should be considered.

Automatic diagnosis of true proximity between the mandibular canal and the third molar on panoramic radiographs using deep learning.

Evaluating the mandibular canal proximity is crucial for planning mandibular third molar extractions. Panoramic radiography is commonly used for radiological examinations before third molar extraction but has limitations in assessing the true contact relationship between the third molars and the mandibular canal. Therefore, the true relationship between the mandibular canal and molars can be determined only through additional cone-beam computed tomography (CBCT) imaging. In this study, we aimed to develop an automatic diagnosis method based on a deep learning model that can determine the true proximity between the mandibular canal and third molars using only panoramic radiographs. A total of 901 third molars shown on panoramic radiographs were examined with CBCT imaging to ascertain whether true proximity existed between the mandibular canal and the third molar by two radiologists (450 molars: true contact, 451 molars: true non-contact). Three deep learning models (RetinaNet, YOLOv3, and EfficientDet) were developed, with performance metrics of accuracy, sensitivity, and specificity. EfficientDet showed the highest performance, with an accuracy of 78.65%, sensitivity of 82.02%, and specificity of 75.28%. The proposed deep learning method can be helpful when clinicians must evaluate the proximity of the mandibular canal and a third molar using only panoramic radiographs without CBCT.

Development and accuracy validation of a fat fraction imaging biomarker for sialadenitis in the parotid gland.

BACKGROUND: The diagnosis of sialadenitis, the most frequent disease of the salivary glands, is challenging when the symptoms are mild. In such cases, biomarkers can be used as definitive diagnostic indicators. Recently, biomarkers have been developed by extracting and analyzing pathological and morphological features from medical imaging. This study aimed to establish a diagnostic reference for sialadenitis based on the quantitative magnetic resonance imaging (MRI) biomarker IDEAL-IQ and assess its accuracy. METHODS: Patients with sialadenitis (n = 46) and control subjects (n = 90) that underwent MRI were selected. Considering that the IDEAL-IQ value is a sensitive fat fractional marker to the body mass index (BMI), all subjects were also categorized as under-, normal-, and overweight. The fat fraction of parotid gland in the control and sialadenitis groups were obtained using IDEAL-IQ map. The values from the subjects in the control and sialadenitis groups were compared in each BMI category. For comparison, t-tests and receiver operating characteristic (ROC) curve analyses were performed. RESULTS: The IDEAL-IQ fat faction of the control and sialadenitis glands were 38.57% and 23.69%, respectively, and the differences were significant. The values were significantly lower in the sialadenitis group (P), regardless of the BMI types. The area under the ROC curve (AUC) was 0.83 (cut-off value: 28.72) in patients with sialadenitis. The AUC for under-, normal-, and overweight individuals were 0.78, 0.81, and 0.92, respectively. CONCLUSIONS: The fat fraction marker based on the IDEAL-IQ method was useful as an objective indicator for diagnosing sialadenitis. This marker would aid less-experienced clinicians in diagnosing sialadenitis.

Radiomics-based sialadenitis staging in contrast-enhanced computed tomography and ultrasonography: a preliminary rat model study.

OBJECTIVE: The aim of this study was to measure the ability of radiomics analysis to diagnose different stages of sialadenitis, compare the diagnostic accuracy of computed tomography (CT) and ultrasonography (US), and suggest radiomics features selected through 3 machine learning algorithms that would be helpful in discriminating between stages of sialadenitis with both imaging systems. STUDY DESIGN: Wistar rats were treated to induce acute and chronic sialadenitis in the left and right submandibular glands, respectively. Contrast-enhanced CT and US of the glands were performed, followed by extirpation and histopathologic confirmation. Radiomics feature values of the glands were obtained from all images. Based on 3 feature selection methods, an optimal feature set was defined after a comparison of the receiver operating characteristic area under the curve (AUC) of each combination of 3 deep learning algorithms and 3 classification models. RESULTS: The attribute features for the CT model were 2 gray-level run length matrices and 2 gray-level zone length matrices. In the US model, there were 2 gray-level co-occurrence matrices and 2 gray-level zone length matrices. The most accurate diagnostic models of CT and US yielded outstanding (AUC = 1.000) and excellent (AUC = 0.879) discrimination, respectively. CONCLUSIONS: The radiomics diagnostic model using gray-level zone length matrices-based features conferred clinically outstanding discriminating ability among stages of sialadenitis using CT and excellent discrimination with US in almost all combinations of machine learning feature selections and classification models.

Multi-planar 2.5D U-Net for image quality enhancement of dental cone-beam CT.

Cone-beam computed tomography (CBCT) can provide 3D images of a targeted area with the advantage of lower dosage than multidetector computed tomography (MDCT; also simply referred to as CT). However, in CBCT, due to the cone-shaped geometry of the X-ray source and the absence of post-patient collimation, the presence of more scattering rays deteriorates the image quality compared with MDCT. CBCT is commonly used in dental clinics, and image artifacts negatively affect the radiology workflow and diagnosis. Studies have attempted to eliminate image artifacts and improve image quality; however, a vast majority of that work sacrificed structural details of the image. The current study presents a novel approach to reduce image artifacts while preserving details and sharpness in the original CBCT image for precise diagnostic purposes. We used MDCT images as reference high-quality images. Pairs of CBCT and MDCT scans were collected retrospectively at a university hospital, followed by co-registration between the CBCT and MDCT images. A contextual loss-optimized multi-planar 2.5D U-Net was proposed. Images corrected using this model were evaluated quantitatively and qualitatively by dental clinicians. The quantitative metrics showed superior quality in output images compared to the original CBCT. In the qualitative evaluation, the generated images presented significantly higher scores for artifacts, noise, resolution, and overall image quality. This proposed novel approach for noise and artifact reduction with sharpness preservation in CBCT suggests the potential of this method for diagnostic imaging.

Development of deep learning model and evaluation in real clinical practice of lingual mandibular bone depression (Stafne cyst) on panoramic radiographs.

OBJECTIVES: Lingual mandibular bone depression (LMBD) is a developmental bony defect in the lingual aspect of the mandible that does not require any surgical treatment. It is sometimes confused with a cyst or another radiolucent pathologic lesion on panoramic radiography. Thus, it is important to differentiate LMBD from true pathological radiolucent lesions requiring treatment. This study aimed to develop a deep learning model for the fully automatic differential diagnosis of LMBD from true pathological radiolucent cysts or tumors on panoramic radiographs without a manual process and evaluate the model's performance using a test dataset that reflected real clinical practice. METHODS: A deep learning model using the EfficientDet algorithm was developed with training and validation data sets (443 images) consisting of 83 LMBD patients and 360 patients with true pathological radiolucent lesions. The test data set (1500 images) consisted of 8 LMBD patients, 53 patients with pathological radiolucent lesions, and 1439 healthy patients based on the clinical prevalence of these conditions in order to simulate real-world conditions, and the model was evaluated in terms of accuracy, sensitivity, and specificity using this test data set. RESULTS: The model's accuracy, sensitivity, and specificity were more than 99.8%, and only 10 out of 1500 test images were erroneously predicted. CONCLUSION: Excellent performance was found for the proposed model, in which the number of patients in each group was composed to reflect the prevalence in real-world clinical practice. The model can help dental clinicians make accurate diagnoses and avoid unnecessary examinations in real clinical settings.

Refinement of image quality in panoramic radiography using a generative adversarial network.

OBJECTIVE: We aimed to develop and assess the clinical usefulness of a generative adversarial network (GAN) model for improving image quality in panoramic radiography. METHODS: Panoramic radiographs obtained at Yonsei University Dental Hospital were randomly selected for study inclusion (n = 100). Datasets with degraded image quality (n = 400) were prepared using four different processing methods: blur, noise, blur with noise, and blur in the anterior teeth region. The images were distributed to the training and test datasets in a ratio of 9:1 for each group. The Pix2Pix GAN model was trained using pairs of the original and degraded image datasets for 100 epochs. The peak signal-to-noise ratio (PSNR) and structural similarity index measure (SSIM) were obtained for the test dataset, and two oral and maxillofacial radiologists rated the quality of clinical images. RESULTS: Among the degraded images, the GAN model enabled the greatest improvement in those with blur in the region of the anterior teeth but was least effective in improving images exhibiting blur with noise (PSNR, 36.27 > 32.74; SSIM, 0.90 > 0.82). While the mean clinical image quality score of the original radiographs was 44.6 out of 46.0, the highest and lowest predicted scores were observed in the blur (45.2) and noise (36.0) groups. CONCLUSION: The GAN model developed in this study has the potential to improve panoramic radiographs with degraded image quality, both quantitatively and qualitatively. As the model performs better in refining blurred images, further research is required to identify the most effective methods for handling noisy images.

Deep learning synthesis of cone-beam computed tomography from zero echo time magnetic resonance imaging.

Cone-beam computed tomography (CBCT) produces high-resolution of hard tissue even in small voxel size, but the process is associated with radiation exposure and poor soft tissue imaging. Thus, we synthesized a CBCT image from the magnetic resonance imaging (MRI), using deep learning and to assess its clinical accuracy. We collected patients who underwent both CBCT and MRI simultaneously in our institution (Seoul). MRI data were registered with CBCT data, and both data were prepared into 512 slices of axial, sagittal, and coronal sections. A deep learning-based synthesis model was trained and the output data were evaluated by comparing the original and synthetic CBCT (syCBCT). According to expert evaluation, syCBCT images showed better performance in terms of artifacts and noise criteria but had poor resolution compared to the original CBCT images. In syCBCT, hard tissue showed better clarity with significantly different MAE and SSIM. This study result would be a basis for replacing CBCT with non-radiation imaging that would be helpful for patients planning to undergo both MRI and CBCT.

Korean dental hygiene students' perceptions and attitudes toward artificial intelligence: An online survey.

OBJECTIVES: This study investigated Korean dental hygiene students' perceptions and attitudes toward artificial intelligence (AI) and aimed to identify needs for education to strengthen professional competencies. METHODS: A 24-question online survey was conducted to the dental hygiene students from four Korean schools in 2021. The questionnaire included seven questions on basic characteristics and 17 AI-related questions on the student's attitudes toward AI, the confidence in AI, predictions about AI, and its future prospects. Responses were analyzed according to the frequencies and correlations between the participants' subjective level of knowledge about AI and questions using chi-square test. RESULTS: Invitations were sent out to 1310 students and 800 (61.1%) participated. Note that 44.2% of participants were interested in AI, and 93.1% accessed AI-related information through the internet. Participants expressed lower confidence in AI's diagnosis (14.8%) and judgment (8.1%) than in those of humans, and 21.9% believed AI would replace their job. The proportions of participants with positive perceptions of the usefulness and the potential for improvement of AI in dentistry were 65.5% and 55.4%, respectively. Participants from schools who had existing AI knowledge expressed higher demands for AI-related content as compared to those who did not (p < 0.05). CONCLUSION: Although dental hygiene students expressed low level of confidence in AI, they were interested in AI and had positive views of its application and potential for improvement. However, the fact they had little AI-related information from dental hygiene curriculum strongly suggests the need for AI-related lectures in schools to prepare for the future.

Automatic diagnosis of retention pseudocyst in the maxillary sinus on panoramic radiographs using a convolutional neural network algorithm.

The evaluation of the maxillary sinus is very important in dental practice such as tooth extraction and implantation because of its proximity to the teeth, but it is not easy to evaluate because of the overlapping structures such as the maxilla and the zygoma on panoramic radiographs. When doom-shaped retention pseudocysts are observed in sinus on panoramic radiographs, they are often misdiagnosed as cysts or tumors, and additional computed tomography is performed, resulting in unnecessary radiation exposure and cost. The purpose of this study was to develop a deep learning model that automatically classifies retention pseudocysts in the maxillary sinuses on panoramic radiographs. A total of 426 maxillary sinuses from panoramic radiographs of 213 patients were included in this study. These maxillary sinuses included 86 sinuses with retention pseudocysts, 261 healthy sinuses, and 79 sinuses with cysts or tumors. An EfficientDet model first introduced by Tan for detecting and classifying the maxillary sinuses was developed. The developed model was trained for 200 times on the training and validation datasets (342 sinuses), and the model performance was evaluated in terms of accuracy, sensitivity, and specificity on the test dataset (21 retention pseudocysts, 43 healthy sinuses, and 20 cysts or tumors). The accuracy of the model for classifying retention pseudocysts was 81%, and the model also showed higher accuracy for classifying healthy sinuses and cysts or tumors (98% and 90%, respectively). One of the 21 retention pseudocysts in the test dataset was misdiagnosed as a cyst or tumor. The proposed model for automatically classifying retention pseudocysts in the maxillary sinuses on panoramic radiographs showed excellent diagnostic performance. This model could help clinicians automatically diagnose the maxillary sinuses on panoramic radiographs.

Impact of physiological parameters on the parotid gland fat fraction in a normal population.

Quantifying physiological fat tissue in the organs is important to further assess the organ's pathologic status. This study aimed to investigate the impact of body mass index (BMI), age, and sex on the fat fraction of normal parotid glands. Patients undergoing magnetic resonance imaging (MRI) of iterative decomposition of water and fat with echo asymmetry and least squares estimation (IDEAL-IQ) due to non-salivary gland-related disease were reviewed. Clinical information of individual patients was categorized into groups based on BMI (under/normal/overweight), age (age I/age II/age III), and sex (female/male) and an inter-group comparison of the fat fraction values of both parotid glands was conducted. Overall, in the 626 parotid glands analyzed, the fat fraction of the gland was 35.80%. The mean fat fraction value increased with BMI (30.23%, 35.74%, and 46.61% in the underweight, normal and overweight groups, respectively [p < 0.01]) and age (32.42%, 36.20%, and 41.94% in the age I, II, and III groups, respectively [p < 0.01]). The fat content of normal parotid glands varies significantly depending on the body mass and age regardless of sex. Therefore, the patient's age and body mass should be considered when evaluating fatty change in the parotid glands in imaging results.

Parotid gland evaluation of menopausal women with xerostomia using the iterative decomposition of water and fat with echo asymmetry and least-squares estimation (IDEAL-IQ) method of MRI: a pilot study.

OBJECTIVES: This study aimed to analyze the quantitative fat fraction (FF) of the parotid gland in menopausal females with xerostomia using the iterative decomposition of water and fat with echo asymmetry and least-squares estimation (IDEAL-IQ) method. METHODS: A total 138 parotid glands of 69 menopausal females were enrolled in our study and participants were divided into normal group and xerostomia group. The xerostomia group was divided into those with or without Sjögren's syndrome. Participants underwent IDEAL-IQ sequences of MRI and the stimulated salivary flow test (s-SFR). The unpaired t-test was used to compare the FFs between the normal and xerostomia groups and between the subgroups with and without Sjögren's syndrome. The correlation between FF and s-SFR was analyzed by Pearson's correlation. RESULTS: Excellent intra- and interobserver agreement during the measurement of FFs by IDEAL-IQ method (ICC>0.99, respectively). FF value in the xerostomia group was statistically significantly higher than the value in the normal group (p < 0.05). Within the xerostomia group, the average FF value of females with Sjögren's syndrome was higher than that of females without Sjögren's syndrome. However, the difference was not statistically significant (p > 0.05). Within the xerostomia group, FF value correlated negatively with s-SFR (p < 0.05). CONCLUSIONS: The FF of the parotid gland was higher in the xerostomia group than in the normal group and FF value and s-SFR showed a negative correlation. Analyses of the FF using IDEAL-IQ in menopausal females can be helpful for the quantitative diagnosis of xerostomia.

Radiomics approach to the condylar head for legal age classification using cone-beam computed tomography: A pilot study.

Legal age estimation of living individuals is a critically important issue, and radiomics is an emerging research field that extracts quantitative data from medical images. However, no reports have proposed age-related radiomics features of the condylar head or an age classification model using those features. This study aimed to introduce a radiomics approach for various classifications of legal age (18, 19, 20, and 21 years old) based on cone-beam computed tomography (CBCT) images of the mandibular condylar head, and to evaluate the usefulness of the radiomics features selected by machine learning models as imaging biomarkers. CBCT images from 85 subjects were divided into eight age groups for four legal age classifications: ≤17 and ≥18 years old groups (18-year age classification), ≤18 and ≥19 years old groups (19-year age classification), ≤19 and ≥20 years old groups (20-year age classification) and ≤20 and ≥21 years old groups (21-year age classification). The condylar heads were manually segmented by an expert. In total, 127 radiomics features were extracted from the segmented area of each condylar head. The random forest (RF) method was utilized to select features and develop the age classification model for four legal ages. After sorting features in descending order of importance, the top 10 extracted features were used. The 21-year age classification model showed the best performance, with an accuracy of 91.18%, sensitivity of 80%, and specificity of 95.83%. Radiomics features of the condylar head using CBCT showed the possibility of age estimation, and the selected features were useful as imaging biomarkers.

Comparison of the clinical usefulness of structured and free-text reports for interpretation of jaw lesions on cone beam computed tomography images.

OBJECTIVE: This study compared the clinical usefulness of structured reports (SRs) and free-text reports (FTRs) of lesions depicted on cone beam computed tomography (CBCT) images from the perspectives of report providers and receivers. STUDY DESIGN: In total, 36 CBCT images of jaw lesions obtained between February 2020 and August 2020 were evaluated. A working group of 3 oral and maxillofacial radiologists (OMRs) established a reporting system and prepared reports. Evaluation group I (2 OMRs) wrote SRs and FTRs for each case and assessed the reporting process for the criteria of convenience and organization. Evaluation group II (3 general practitioners [GPs] and 3 oral and maxillofacial surgeons [OMSs]) assessed the reports for the criteria of productivity, consistency, and organization. A 5-point Likert scale was used to assess the usefulness of each report. Scores were statistically compared according to report type with the paired Wilcoxon signed-rank test. RESULTS: The SRs scored significantly higher for all criteria as assessed by evaluation group I and the GPs of group II (P < .001). The FTRs scored significantly higher for productivity and organization as assessed by the OMSs of group II (P = .005 for both criteria). CONCLUSIONS: The clinical usefulness of reports may differ according to roles of the report recipients in diagnosis and treatment.

Harmonization of robust radiomic features in the submandibular gland using multi-ultrasound systems: a preliminary study.

OBJECTIVE: This study aimed to identify robust radiomic features in multiultrasonography of the submandibular gland and normalize the interdevice discrepancies by applying a machine-learning-based harmonization method. METHODS: Ultrasonographic images of normal submandibular gland of young healthy adults, aged between 20 and 40 years, were selected from two different devices. In a total of 30 images, the region of interest was determined along the border of gland parenchyma, and 103 radiomic features were extracted using A-VIEW. The coefficient of variation (CV) was obtained for individual features, and the features showing CV less than 10% were selected. For the selected features, the interdevice discrepancy was normalized using machine-learning method, called the ComBat harmonization. Median differences of the features between the two scanners, before and after harmonization, were compared using Mann-Whitney U-test; confidence interval of 95%. RESULTS: Among total 103 radiomic features, 17 features were selected as robust, showing CV less than 10% in both scanners. All values of selected features, except two, showed a statistical difference between the two devices. After applying the ComBat harmonization method, the median and distribution of the 16 features were harmonized to show no significant difference between the two scanners (p > 0.05). One feature remained different (p ≤ 0.05). CONCLUSION: On ultrasonographic examination, robust radiomic features for normal submandibular gland were obtained and interdevice normalization was efficiently conducted using ComBat harmonization. Our findings would be useful for multidevices or multicenter studies based on clinical ultrasonographic imaging data to improve the accuracy of the overall diagnostic model.

Performance comparison of three deep learning models for impacted mesiodens detection on periapical radiographs.

This study aimed to develop deep learning models that automatically detect impacted mesiodens on periapical radiographs of primary and mixed dentition using the YOLOv3, RetinaNet, and EfficientDet-D3 algorithms and to compare their performance. Periapical radiographs of 600 pediatric patients (age range, 3-13 years) with mesiodens were used as a training and validation dataset. Deep learning models based on the YOLOv3, RetinaNet, and EfficientDet-D3 algorithms for detecting mesiodens were developed, and each model was trained 300 times using training (540 images) and validation datasets (60 images). The performance of each model was evaluated based on accuracy, sensitivity, and specificity using 120 test images (60 periapical radiographs with mesiodens and 60 periapical radiographs without mesiodens). The accuracy of the YOLOv3, RetinaNet, and EfficientDet-D3 models was 97.5%, 98.3%, and 99.2%, respectively. The sensitivity was 100% for both the YOLOv3 and RetinaNet models and 98.3% for the EfficientDet-D3 model. The specificity was 100%, 96.7%, and 95.0% for the EfficientDet-D3, RetinaNet, and YOLOv3 models, respectively. The proposed models using three deep learning algorithms to detect mesiodens on periapical radiographs showed good performance. The EfficientDet-D3 model showed the highest accuracy for detecting mesiodens on periapical radiographs.