Imaging Science in Dentistry最新文献

Purpose: The aim of this study was to assess radiomorphometric indices in panoramic radiographs of pediatric cancer survivors.

Materials and methods: This population-based case-control study included 31 patients who were 18 years or older at the time of the radiograph and had a history of cancer diagnosed during childhood or adolescence (between 0 and 19 years of age), treated with chemotherapy and/or radiotherapy. The radiomorphometric indices mandibular cortical width (MCW) and mandibular cortical index (MCI) were assessed in panoramic radiographs to determine whether there was evidence of a reduction in bone mineral density (BMD). A 95% confidence level (P<0.05) was used for all statistical tests.

Results: In the MCW analysis, the case group showed a variation between 2.2 mm and 4.6 mm, while the control group showed a variation between 2.9 mm and 5.2 mm (P<0.05). For MCI, most (54.8%) of the images in the case group were classified as C2, and 16.1% as C3, whereas none in the control group were classified as C3, demonstrating a statistically significant difference (P<0. 05).

Conclusion: Childhood cancer survivors are at an increased risk of long-term reduction in BMD, and panoramic radiomorphometric indices represent an accessible and reliable screening tool for predicting this risk.

Purpose: This study aimed to compare the diagnostic performance of cone-beam computed tomography (CBCT) and digital periapical radiography in detecting separated endodontic instruments (SI) according to their type and length in filled root canals of mandibular molars.

Materials and methods: Forty-two extracted mandibular molars were divided into 2 groups: control (without SI) and experimental (with SI). In the experimental group, SI fragments (2 or 3 mm) from 3 endodontic instruments (WaveOne Gold, Reciproc Blue, and ProTaper Next) were inserted into the mesiobuccal canals of 30 teeth. CBCT scans were obtained using the KAVO OP 3D PRO device, and digital periapical radiographs (orthoradial, distal, and mesial) were acquired using the Focus X-ray unit with the RVG 5200 solid sensor. Three experts independently evaluated all images for SI detection using a 5-point scale. Diagnostic performance metrics were calculated for each imaging method, and comparisons were performed using 1-way analysis of variance with Tukey post-hoc tests. The significance level was set at 5% (P=0.05).

Results: Digital periapical radiography demonstrated higher area under the receiver operating characteristic curve (Az), sensitivity, and specificity values than CBCT. ProTaper Next SI on CBCT scans showed the lowest Az values (0.58 and 0.56 for 2- and 3-mm fragments, respectively). These values were significantly different from those of the other SI conditions (both CBCT and digital periapical radiography) (P<0.05).

Conclusion: Digital periapical radiography outperformed CBCT in detecting SI in filled root canals of mandibular molars and should be the preferred modality. Furthermore, SI detectability varies with imaging modality, depending on instrument type and fragment length.

Purpose: The objective of this study was to evaluate variations in entrance skin dose across different anatomical regions of a head phantom and to present entrance skin dose results obtained from a limited number of measurement points. By identifying key anatomical landmarks, the study aims to rationalize the number of entrance skin dose measurement points using thermoluminescence dosimeters (TLDs).

Materials and methods: A tissue-equivalent head and neck phantom was examined with a 1-shot cephalometric X-ray machine under standardized settings. A detailed measurement protocol (140 TLDs for protocol 1) was compared to 3 alternative protocols using fewer measurement points (40 TLDs for protocol 2, 30 TLDs for protocol 3, and 6 TLDs for protocol 4).

Results: Protocol 1 yielded an average entrance skin dose of 1.042 mSv, with higher doses at the parotid gland and above the left eye corner. Protocols 2 and 3 produced similar mean doses (1.003 mSv and 1.04 mSv, respectively), demonstrating that fewer dosimeters can still generate reliable results. Protocol 4, which targeted radiosensitive sites such as the thyroid and parotid glands, provided a comparable mean entrance skin dose for the beam-facing side (1.76 mSv) relative to the other protocols.

Conclusion: These findings suggest that the selected dosimetry approach is effective in assessing exposure to key tissues such as the skin. The results provide a foundation for a more representative, reproducible, and cost-effective dosimetric method in cephalometric radiography.

Purpose: This study assessed the effects of cone-beam computed tomography (CBCT) voxel size and exposure settings on the accuracy of alveolar bone thickness (BT) measurements.

Materials and methods: Seven pig mandibles were scanned using 6 CBCT settings: 200, 400, and 600 µm voxels, each with an ultra-low dose (ULD) and a standard dose (SD) setting. BT was measured at 3 locations on 1 anterior and 1 posterior tooth. These teeth were then scanned using micro-computed tomography (micro-CT), and BT was measured at the same locations. The intraclass correlation coefficient (ICC) and paired t-test were used to evaluate agreement between CBCT and micro-CT measurements. The discrepancy between CBCT and micro-CT measurements was calculated, and repeated-measures analysis of variance (RMANOVA) was used to evaluate the effects of the 6 CBCT settings on BT measurement accuracy.

Results: The lowest discrepancy between micro-CT and CBCT was observed with the 600 µm ULD (0.29±0.39 mm), followed by the 200 µm SD (0.41±0.72 mm). When BT was less than 2 mm, the lowest discrepancy occurred with the 200 µm SD (0.29±0.33 mm), followed by the 600 µm ULD (0.32±0.25 mm). The ICC between CBCT and micro-CT decreased as BT decreased. RMANOVA indicated no significant effect of CBCT settings on measurement accuracy.

Conclusion: BT measured with the 600 µm ULD was comparable to that obtained with the 200 µm SD. The discrepancy between CBCT and micro-CT was notable, especially when BT was less than 2 mm.

Purpose: The aim of this study was to evaluate the effect of 4 different post materials on the contrast-to-noise ratio (CNR) in various tooth regions using multiple cone-beam computed tomography (CBCT) imaging settings and metal artifact reduction (MAR) algorithms.

Materials and methods: Forty single-rooted teeth with 4 different post materials were mounted in a skull and scanned using the Planmeca ProMax 3D Max under different conditions: 90 and 96 kVp; no, medium, and high MAR settings; and voxel sizes of 100 and 200 μm. Image analysis was performed with ImageJ software to calculate CNR. The dose-area product (DAP) was recorded, and the effective dose was calculated.

Results: When all other parameters were held constant, each factor significantly influenced mean CNR (P<0.05). Higher CNR values were observed in the middle third of roots at 96 kVp, with a voxel size of 200 μm, and with medium or high MAR. No significant difference was found between medium and high MAR. Gold posts showed the lowest mean CNR, while no significant differences were observed among the other post materials.

Conclusion: The combination of low resolution, medium MAR, and 96 kVp provided the highest image quality and CNR for visualizing teeth with post materials using the Planmeca ProMax, while also producing one of the lowest effective doses.

Purpose: This study investigated the effect of increasing slice thickness and applying sharpening filters on the diagnosis of bone graft loss in regions adjacent to zirconia implants using cone-beam computed tomography (CBCT).

Materials and methods: Twelve zirconia implants were inserted into mandibles, followed by bone graft application. Graft loss was then simulated in half of the samples. CBCT scans were obtained using 2 devices (OP300 and Eagle 3D) with parameters set at 90 kVp, 8 mA, and a 133 µm voxel size. Five examiners assessed the scans under different conditions: original thickness, 1 mm, 2 mm, no filter, Sharpen 1×, and Sharpen 2×. The area under the receiver operating characteristic curve (AUC), sensitivity, and specificity were calculated and compared using 2-way analysis of variance (α=5%).

Results: For the OP300 device, increasing slice thickness to 1 mm impaired AUC, sensitivity, and specificity metrics (P<0.05). Overall, filter application reduced this impairment (P<0.05). For the Eagle device, slice thickness did not affect diagnostic metrics, but specificity values increased with sharpening filter application on scans at original thickness or with 1 mm slices (P<0.05).

Conclusion: For 1 device, a 1 mm increase in slice thickness impaired the diagnosis of bone graft loss in areas adjacent to zirconia implants. However, the application of sharpening filters helped mitigate this impairment.

Purpose: Implant identification is a pressing concern in dental implantology, and artificial intelligence (AI) has been evaluated for this purpose. YOLO, a state-of-the-art object detection model, is suitable for medical imaging; therefore, this study assessed YOLOv11-the latest iteration-for identifying 10 implant types in Indian clinical settings and compared its accuracy to that of dental professionals.

Materials and methods: A dataset of 3,161 radiographs, comprising both periapical and panoramic images of 10 implant types, was annotated and used to train and test YOLOv11. Training was performed on Google Colab using an NVIDIA Tesla T4 GPU (16 GB VRAM). A random sample of 200 radiographs was selected from the test dataset and presented to 50 dental practitioners for implant identification. Their responses were analysed and compared, using the chi-square test for statistical significance.

Results: YOLOv11 achieved precision of 0.87, recall of 0.85, an F1-score of 0.86, and an mAP50 of 0.899. The model achieved excellent classification accuracy for Adin (95%), MIS (94%), Bego (92%), ITI (96%), and Bicon (97%). Moderate accuracy was noted for Noris (82%), Osstem (85%), AlphaBio (88%), Dentium (77%), and Bioline (75%). YOLOv11 demonstrated higher overall accuracy and consistency than dental professionals. Dentists' accuracy ranged from 27% to 49%, whereas that of YOLOv11 ranged from 92% to 100%.

Conclusion: YOLOv11 recognised most implant classes with over 90% accuracy, surpassing traditional manual techniques in implant detection. Although the model is dependable and efficient, certain aspects require improvement. The study also emphasises the significance of a region-specific approach for clinical relevance.

Purpose: This study aimed to develop a deep learning model incorporating the pulp-to-tooth area ratio (PTR) for estimating age from periapical radiographs of upper central incisors in a Thai population.

Materials and methods: A total of 2,041 periapical radiographs, representing 3,108 upper central incisors from individuals aged 10.00 to 84.25 years, were analyzed. Root and root canal segmentation masks were generated using Labelbox. The dataset was randomly divided into training (2,175 teeth), validation (466 teeth), and test (467 teeth) sets. Model development proceeded in 3 steps: (1) DeepLabv3+ was trained to segment root and root canal regions. (2) ResNet-50 was trained to estimate age using both segmented images and PTR values derived from the masks. (3) The validation set guided model selection, and the final model was evaluated on the test set using mean absolute error (MAE).

Results: The segmentation model (DeepLabv3+) achieved a mean intersection over union (mIoU) of 81.85%. The age estimation model (ResNet-50) yielded an overall MAE of 6.14 years. For age-specific evaluation, test data were grouped into 4 categories: 10-19 years, 20-39 years, 40-59 years, and ≥60 years. The MAE was lowest in the 10-19 years group (3.74 years) and progressively increased with age: 6.07, 6.58, and 8.40 years, respectively.

Conclusion: This study demonstrated that integrating deep learning models with PTR measurement offers a promising method for dental radiographic age estimation using a single tooth.

Purpose: This study evaluated how restorative materials-specifically, lithium disilicate (LD) (IPS e.max; Ivoclar Vivadent, Liechtenstein, Switzerland) and zirconia-affect the contrast-to-noise ratio (CNR) in cone-beam computed tomography (CBCT) imaging.

Materials and methods: CNR was compared between LD and zirconia (NexxZr T; Sagemax Bioceramics, Federal Way, USA; and Zirlux Transitions; Zahn Dental, New York, USA) using 2 CBCT machines-Planmeca ProMax 3D Mid 2015 (Planmeca, Helsinki, Finland) and PreXion 3D Excelsior (PreXion Inc., San Mateo, USA)-under multiple exposure settings, with and without metal artifact reduction (MAR). Four cylindrical ingots were scanned in a saline phantom under 10 conditions (Planmeca: Low, Normal, High, ±MAR; PreXion: Rapid, Standard, HD, UHD). CNR was calculated in ImageJ (National Institutes of Health, Bethesda, MD, USA) and analyzed with the paired t-test.

Results: LD showed the highest CNR on Planmeca without MAR and across all PreXion settings. With MAR on Planmeca, zirconia outperformed LD. For zirconia, CNR improved with larger voxel sizes and MAR, whereas for LD it was optimal with smaller voxel sizes and MAR. Optimal Planmeca settings were Low (400 µm)+MAR for zirconia and High (150 µm)+MAR for LD. On PreXion, Rapid (200 µm) favored zirconia, while HD (100 µm) favored LD.

Conclusion: CBCT settings should be tailored to the restorative material, diagnostic need, and radiation dose. For zirconia, larger voxel sizes with MAR enhance CNR; for LD, smaller voxel sizes with MAR are optimal. These findings support imaging optimization for patients with dental restorations.