Dento maxillo facial radiology最新文献

Objectives: This study aimed to systematically review the current performance of large language models (LLMs) in dento-maxillofacial radiology (DMFR).

Methods: Five electronic databases were used to identify studies that developed, fine-tuned, or evaluated LLMs for DMFR-related tasks. Data extracted included study purpose, LLM type, images/text source, applied language, dataset characteristics, input and output, performance outcomes, evaluation methods, and reference standards. Customized assessment criteria adapted from the TRIPOD-LLM reporting guideline were used to evaluate the risk-of-bias in the included studies specifically regarding the clarity of dataset origin, the robustness of performance evaluation methods, and the validity of the reference standards.

Results: The initial search yielded 1621 titles, and 19 studies were included. These studies investigated the use of LLMs for tasks including the production and answering of DMFR-related qualification exams and educational questions (n = 8), diagnosis and treatment recommendations (n = 7), and radiology report generation and patient communication (n = 4). LLMs demonstrated varied performance in diagnosing dental conditions, with accuracy ranging from 37% to 92.5% and expert ratings for differential diagnosis and treatment planning between 3.6 and 4.7 on a 5-point scale. For DMFR-related qualification exams and board-style questions, LLMs achieved correctness rates between 33.3% and 86.1%. Automated radiology report generation showed moderate performance with accuracy ranging from 70.4% to 81.3%.

Conclusions: LLMs demonstrate promising potential in DMFR, particularly for diagnostic, educational, and report generation tasks. However, their current accuracy, completeness, and consistency remain variable. Further development, validation, and standardization are needed before LLMs can be reliably integrated as supportive tools in clinical workflows and educational settings.

Objectives: To investigate the clinical efficiency of an artificial intelligence-based metal artefact correction algorithm (AI-MAC) for reducing dental metal artefacts in head and neck CT, compared to conventional interpolation-based metal artefact correction (MAC).

Methods: We retrospectively collected 41 patients with non-removal dental hardware who underwent non-contrast head and neck CT prior to radiotherapy. All images were reconstructed with the standard reconstruction algorithm (SRA) and were additionally processed with both conventional MAC and AI-MAC. The image quality of SRA, MAC, and AI-MAC was compared by qualitative scoring on a 5-point scale, with scores ≥ 3 considered interpretable. This was followed by a quantitative evaluation, including signal-to-noise ratio (SNR) and artefact index (Idxartefact). Organ contouring accuracy was quantified via calculating the dice similarity coefficient (DSC) and hausdorff distance (HD) for the oral cavity and teeth, using the clinically accepted contouring as reference. Moreover, the treatment planning dose distribution for the oral cavity was assessed.

Results: AI-MAC yielded superior qualitative image quality as well as quantitative metrics, including SNR and Idxartefact, to SRA and MAC. The image interpretability significantly improved from 41.46% for SRA and 56.10% for MAC to 92.68% for AI-MAC (P < .05). Compared to SRA and MAC, the best DSC and HD for both oral cavity and teeth were obtained on AI-MAC (all P < .05). No significant differences for dose distribution were found among the 3 image sets.

Conclusion: AI-MAC outperforms conventional MAC in metal artefact reduction, achieving superior image quality with high image interpretability for patients with dental hardware undergoing head and neck CT. Furthermore, the use of AI-MAC improves the accuracy of organ contouring while providing consistent dose calculation against metal artefacts in radiotherapy.

Advances in knowledge: AI-MAC is a novel deep learning-based technique for reducing metal artefacts on CT. This in vivo study demonstrated its capability of reducing metal artefacts while preserving organ visualization, as compared with conventional MAC.

Objectives: This study evaluated the influence of cognitive aids, including machine learning (ML) algorithms and checklists, on the diagnostic accuracy and confidence of dental students in detecting dental caries on bitewing radiographs.

Methods: Fifty-two third-year dental students were randomly assigned to control, ML, or checklist groups. The participants recorded their caries diagnoses (charting) on 10 bitewing radiographs and rated their confidence. Diagnostic accuracy and reliability were compared between groups for caries detection (present/absent). The inter-rater reliability for International Caries Detection and Assessment System II (ICDAS II) caries grading was assessed using weighted kappa. Participants also completed questionnaires on their perceptions of cognitive aids.

Results: ML group showed the highest diagnostic accuracy and confidence levels. For caries detection, the ML group achieved the highest sensitivity (79%) and diagnostic odds ratio (20.3), while the checklist group had the highest specificity (90.9%) (P < .001). The control group showed moderate sensitivity (67.9%) but outperformed the checklist group in this metric. Inter-rater agreement for caries detection was highest in the ML group (κ = 0.702, 95% CI: 0.692-0.713; P < .001), followed by the checklist group. The ML group also had the highest weighted kappa for ICDAS II grading (κ = 0.520, P < .001). Confidence levels differed significantly between groups (P < .001), with the ML group reporting the highest confidence.

Conclusion: ML algorithms may enhance diagnostic accuracy and confidence, possibly by reducing cognitive load. While standardizing the diagnostic process, checklists were less effective, likely due to their lack of flexibility and clinical context. Further research is needed to better understand our findings and translate them into clinical workflows.

Objectives: This research focuses on detecting the orthogonal plane to the jaw and the dental arch curve on this plane so that high-quality panoramic images can be reconstructed from cone beam CT images.

Methods: The movement trajectory of panoramic reconstruction, known as the dental arch curve, determines the quality of the reconstructed panoramic image. In traditional methods, the dental arch curve is detected on 1 transversal slice using interpolation methods. However, these methods may fail to capture the actual dental arch since the jaw is not usually perpendicular to the transversal slice and the interpolation methods tend to ignore local anatomical information of the jaw and teeth. To improve detection of the actual dental arch, we define the jaw orthogonal plane by establishing a relationship between the plane variables and the jaw's left-to-right and the anterior-and-posterior tilts. On this plane, the dental arch curve is first initialized and then moved to match the actual dental arch.

Results: Experimental results demonstrate that our method accurately detects dental arch curves. Panoramic images reconstructed using these curves effectively display the true anatomical features of the jaw and teeth with less distortion compared to those reconstructed by traditional methods.

Conclusions: Our detected dental arch curve on the jaw orthogonal plane is more accurately located in the middle of the jaw and teeth. Anatomic information of the jaw and teeth around the detected dental curve is rightly employed to reconstruct high-quality panoramic images.

Objectives: To investigate the correlation between elastography values, age, and duration of bruxism by quantitatively measuring masseter muscle (MM) stiffness with shear wave elastography (SWE) in adolescents with bruxism.

Methods: This prospective study evaluated 132 MMs from 66 adolescents: 33 controls and 33 with bruxism. The thickness and stiffness of the MM were measured. The SWE values (metres/second; m/s) and kilopascals (kPa) of the patient and control groups were quantitatively compared.

Results: The elastic and velocity values of the MM in both closed and open positions were higher in bruxism patients compared to controls (P < .001, for each). No significant difference existed in MM thickness (P = .904). The receiver operating characteristic analysis for different SWE values found a sensitivity and specificity at baseline of 0.81 kPa, 0.60 m/s and 0.76 kPa, 0.67 m/s with the mouth closed. The values found with the mouth open were 0.76 kPa, 0.64 m/s and 0.76 kPa, 0.61 m/s.

Conclusions: Adolescents with bruxism had higher MM hardness values in closed and open positions compared to the control group. SWE can be used as an effective imaging method to measure MM hardness. No relationship existed between SWE values and the patient's age or duration of bruxism.

Objective: The aim of this study was to evaluate the impact of a 3D-printed model with simulated oral changes on the teaching of radiographic evaluation.

Method: A model of an adult patient with several simulated alterations was designed, including impacted teeth, dentigerous cyst, mesiodens, coronal fractures, periodontal resorptions, periapical lesions, and exostoses. The radiographic images obtained were evaluated by postgraduate students using a questionnaire. The data obtained were analysed with descriptive and inferential statistics.

Results: The 3D model produced satisfactory images for the simulation of the proposed alterations. The general perception of the participants was positive, but there were significant differences between master's and doctoral students regarding the clinical-radiographic relationship of the simulated changes in general (P = .037) and the radiographic image of impacted canine (P = .032).

Conclusions: The 3D model was positively evaluated in most of the simulated alterations, demonstrating its potential as a pedagogical tool. These results reinforce the feasibility of 3D printing for producing models for radiographic assessment, offering image quality and versatility for the development of complex training.

Advances in knowledge: This is the first study to develop and evaluate a 3D-printed model with complex anatomical and pathological alterations for preclinical training in Oral and Maxillofacial Radiology.

Objective: The aim of this study was to investigate MR sialography and SPECT/CT for parotid glands in Sjögren's syndrome patients.

Methods: Thirty Sjögren's syndrome patients underwent MR sialography and SPECT/CT. The MR sialographic stagings of Sjögren's syndrome were determined by the criteria (stage 0: normal; stage 1: punctate appearance; stage 2: globular appearance; stage 3: cavitary appearance; stage 4: destructive appearance). The maximum standardized uptake value (SUVmax) of the right and left parotid glands with SPECT/CT was obtained using a workstation and software. MR sialographic stagings and SUVmax of parotid glands were evaluated at pre- and post-stimulation and ratio of pre- to post-stimulation.

Results: Regarding pre-stimulation, the SUVmax of stage 0 (31.9 ± 19.3) was significantly higher than that of stage 2 (19.7 ± 7.5, P = .046), stage 3 (10.2 ± 7.1, P < .001) and stage 4 (6.8 ± 4.3, P < .001). Furthermore, the SUVmax at ratio of pre- to post-stimulation of stage 0 (1.87 ± 0.55) was significantly higher than that of stage 3 (1.16 ± 0.30, P = .001) and stage 4 (1.16 ± 0.40, P < .001).

Conclusion: This study suggests that MR sialography and SPECT/CT SUV are effective tool for the management of Sjögren's syndrome patients.

Objectives: To assess the influence of disinfecting a photostimulable phosphor plate (PSP) receptor with 0.2% peracetic acid on the vertical root fracture (VRF) diagnosis.

Methods: Baseline radiographs of 20 single-rooted teeth (10 without VRF and 10 with VRF) inserted in an alveolar socket of a human mandible were obtained with an unused PSP receptor of the Express digital system (Instrumentarium Dental Inc., Milwaukee, United States) and a Focus X-ray unit (Instrumentarium, Tuusula, Finland) set at 70 kVp, 7 mA, and an exposure time of 0.125 s. Then, 20 disinfections were performed on the PSP receptor, representing one disinfection cycle. Each disinfection lasted 30 s and the interval between them was 40 min. Subsequently, another 20 radiographs were obtained. This process occurred 9 more times, resulting in 200 disinfections (10 cycles × 20 disinfections) and 220 radiographs ([10 cycles × 20 teeth] + 20 baseline radiographs). All 220 radiographs were assessed independently by 5 examiners for VRF diagnosis. Area under the receiver operating characteristic curve (AUC), sensitivity, and specificity were calculated and compared among the number of disinfections by analysis of variance. The significance level was set at 5%. Weighted Kappa test evaluated intra- and inter-examiner agreements.

Results: The disinfections did not affect the AUC, sensitivity and specificity values for VRF diagnosis (P > 0.05). Moreover, the intra- and inter-examiner agreements ranged from moderate to perfect (0.55-1.00) and from fair to moderate (0.22-0.49), respectively.

Conclusions: Disinfecting a PSP receptor with 0.2% peracetic acid did not affect the radiographic diagnosis of VRF.

To investigate the effectiveness of Virtual Reality (VR) simulations in improving learning outcomes in dental radiology in panoramic radiography. Thirty-two volunteer dental students (first and second year of study) had the same theoretical lecture about panoramic imaging. For practical training, students were randomly divided into 2 groups: G1: traditional educational method and G2: VR (Qbion AB software). Panoramic images of 2 anthropomorphic phantoms using a VistaVoxS 3D device were acquired. The type and number of positioning errors and the need for a retake were evaluated by 1 Oral Radiology teacher and 1 senior radiographer in consensus. To test the retention of knowledge, students from both groups had to identify the absence or presence of positioning errors. Data obtained were evaluated by a descriptive analysis and it considered the frequency of the categorical variables. The positioning error rate was higher for G1 (62.50%). Type error 3 (Patient's chin raised too high) was the most prevalent (47.06%) among the groups. The retake rate among all students was similar (25%). G2 was more able to detect patient's positioning errors (68.8%) than G1. The semi-immersive VR software demonstrated potential benefits for dental students. VR tools could be integrated into Oral Radiology preclinical simulations as an additional educational tool to help reduce patient positioning errors in panoramic radiography. This study highlights the effectiveness of semi-immersive VR in improving dental students' ability to detect and prevent positioning errors in panoramic radiography. VR training enhances knowledge retention and supports its integration into preclinical education as an additional educational tool to optimize radiographic training.

Objectives: Reliable cancellous bone segmentation in Cone Beam CT (CBCT) images is essential for post-orthognathic assessment of condylar resorption. However, challenges such as edge blurring and low contrast in CBCT images make effective segmentation difficult. This study aims to overcome these issues, providing a foundation for accurate bone quantification to enhance surgical planning and patient outcomes.

Methods: We propose a novel approach to enhance edge-based segmentation for cancellous bone in CBCT images. By incorporating edge features from the cancellous bone region and utilizing cancellous edge localization as an auxiliary task via Dual-Branch Fusion Network (DBF-Net), our model leverages shared feature parameters across functions to improve segmentation accuracy and robustness.

Results: Our DBF-Net outperformed other models, achieving DICE coefficient of 91.48%. And the 95% Hausdorff Distance decreased to 3.88 mm, demonstrating significant improvement in cancellous bone boundary detection, which is crucial for the post-orthognathic assessment of condylar resorption.

Conclusion: This method provides a robust solution for reliable cancellous bone segmentation in CBCT images to support the quantitative assessment of condylar resorption.