Accuracy of implant height and width measurement with triaxial rotation method based on cone-beam CT.

Objective: To investigate the accuracy of implant height and width measurement in the mandibular and maxillary first molar region based on cone-beam CT (CBCT) data, and to establish an accurate method for bone measurement in the implant region. Materials and methods: CBCT images of 122 patients with implant in mandibular or maxillary first molar region were retrospectively collected. Two methods were used to measure sagittal height (SH), coronal height (CH), sagittal width (SW), and coronal width (CW) of implants. Method 1 (general method): the images were analyzed using the built-in software NNT 9.0 software. SHl, CHl, SWl, and CWl were measured on the reconstructed sagittal and coronal based on the radiologist's own experience. Method 2 (triaxial rotation method): the raw data were demonstrated in Expert mode of NNT 9.0 software, in which the coronal axis and sagittal axis were rotated paralleling to the long axis of the implant for reconstruction, and then SH2, CH2, SW2, and CW2 were measured on the reconstructed sagittal and coronal images. The results of two methods were compared with the actual implant size (H0, W0). Paired T-test was performed for statistical analysis. Dahlberg formula was used to check the measurement error. Results: For method 1, there was no significant differences between SHl and H0 (P > 0.05), but significant differences between CHl and H0, SWl and W0, and CWl and W0 (P < 0.05). For method 2, there were no significant differences between all measurements and actual size (P > 0.05). The random error range measured using Dahlberg formula was 0.157-1.171 mm for general method and 0.017-0.05 mm for triaxial rotation method. Conclusion: The triaxial rotation method is accurate for implant height and width measurements on CBCT images and could be used in pre-operatively bone height and width measurement of potential implant sites.

The exploration of a compound cone-beam CT contrast agent for diagnosis of human extracted cracked tooth.

Objectives: This study aims to investigate the use of sodium iodide (NaI), dimethyl sulfoxide (DMSO), ethyl alcohol, and ethyl acetate as cone-beam CT (CBCT) contrast agents for diagnosing cracked teeth. The optimal delay time for detecting the number of crack lines beyond the dentino-enamel junction (Nd), the number of cracks extending from the occlusal surface to the pulp cavity (Np), and the depth of the crack lines was explored. Methods: 14 human extracted cracked teeth were collected, 12 were used for enhanced scanning, and 2 were used for exploring the characteristic of crack lines. The teeth were scanned in 3 CBCT enhanced scanning (ES) modes: ES1 using meglumine diatrizoate (MD); ES2 using NaI and DMSO, ES3 using NaI, DMSO, ethyl alcohol and ethyl acetate. Three delay times (15mins, 30mins, and 60mins) were set for scanning. Nd, Np, and depth of crack lines were evaluated. Results: There were totally 24 crack lines on 12 cracked teeth. Nd was 10 in ES1 at 60mins, 24 in ES2 at 60mins and 24 in ES3 at 15mins. Np was 1 in ES1 at 60mins, 10 in ES2 at 60mins and 21 in ES3 at 60mins, and there were significantly different among them (p < 0.01). The average depth presented on ES3 was significantly deeper than ES1 and ES2 (p < 0.01). Conclusion: NaI, DMSO, ethyl alcohol and ethyl acetate show potential as contrast agents for enhanced CBCT scanning in diagnosis of cracked teeth and their depth in vivo. A delay time of 15 min is necessary to confirm the existence of crack lines, while a longer delay time is required to ascertain if these crack lines extend to the pulp cavity.

Development of a predictive model for identifying previously undetected vertical root fractures.

This study aimed to develop a predictive model to screen for undetected vertical root fractures (VRFs) in root canal treated teeth. We included 95 root canal treated teeth with suspected VRFs; 77 for training and 18 for validation. Following clinical and cone-beam CT parameters were recorded: sex, tooth type, coronal restoration, time interval from completion of endodontic treatment to definitive diagnosis (TI), type of bone loss (BL), apical extent of root filling (AR) and the ratio of root filling diameter to the actual diameter in the coronal (1/3TA) and middle (2/3TA) root thirds. A predictive model p = 1/(1 - e-x ) was generated, where x = -7.433 + 1.977BL + 1.479 (2/3TA) + 1.102 AR; the sensitivity and specificity were 0.852 and 0.875 for training and 0.917 and 0.833 for validation. VRF teeth were more likely to have vertical bone loss and overfilled root canals. This model had a high diagnostic efficacy for VRFs.

Diagnosis of in vivo vertical root fracture using deep learning on cone-beam CT images.

OBJECTIVES: Evaluating the diagnostic efficiency of deep learning models to diagnose vertical root fracture in vivo on cone-beam CT (CBCT) images. MATERIALS AND METHODS: The CBCT images of 276 teeth (138 VRF teeth and 138 non-VRF teeth) were enrolled and analyzed retrospectively. The diagnostic results of these teeth were confirmed by two chief radiologists. There were two experimental groups: auto-selection group and manual selection group. A total of 552 regions of interest of teeth were cropped in manual selection group and 1118 regions of interest of teeth were cropped in auto-selection group. Three deep learning networks (ResNet50, VGG19 and DenseNet169) were used for diagnosis (3:1 for training and testing). The diagnostic efficiencies (accuracy, sensitivity, specificity, and area under the curve (AUC)) of three networks were calculated in two experiment groups. Meanwhile, 552 teeth images in manual selection group were diagnosed by a radiologist. The diagnostic efficiencies of the three deep learning network models in two experiment groups and the radiologist were calculated. RESULTS: In manual selection group, ResNet50 presented highest accuracy and sensitivity for diagnosing VRF teeth. The accuracy, sensitivity, specificity and AUC was 97.8%, 97.0%, 98.5%, and 0.99, the radiologist presented accuracy, sensitivity, and specificity as 95.3%, 96.4 and 94.2%. In auto-selection group, ResNet50 presented highest accuracy and sensitivity for diagnosing VRF teeth, the accuracy, sensitivity, specificity and AUC was 91.4%, 92.1%, 90.7% and 0.96. CONCLUSION: In manual selection group, ResNet50 presented higher diagnostic efficiency in diagnosis of in vivo VRF teeth than VGG19, DensenNet169 and radiologist with 2 years of experience. In auto-selection group, Resnet50 also presented higher diagnostic efficiency in diagnosis of in vivo VRF teeth than VGG19 and DensenNet169. This makes it a promising auxiliary diagnostic technique to screen for VRF teeth.

Establishment of Narrow Root Fracture Models Using Modified Temperature Cycling Method and Diagnosis Using Different Cone-Beam CT Units.

Aim: Using a modified thermal cycling method to establish narrow root fracture models and evaluate the diagnosis efficiency of them using four different cone-beam CT (CBCT) units. Methodology. Fifty-six intact teeth were selected, and the crowns of the teeth were embedded using general purpose acrylic resin. 50 root fracture models were established by soaking these teeth in liquid nitrogen and hot water cyclically; 6 teeth were used as the negative control. All the 56 teeth were scanned with the smallest voxel size of four different CBCT units (NewTom VGi, Planmeca Promax 3D Max, Kavo 3D eXam, and Soredex Scanora3D). 10 teeth were randomly selected, and the roots were sliced using slow-speed saw to obtain horizontal root sections. Scanning electron microscope (SEM) was used to measure the width of the fracture lines (FLs). The CBCT images were evaluated for the presence or absence of fracture lines. Accuracy, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated for the diagnosis of FLs using the four CBCT units. Results: Fifty narrow root fracture models were successfully established, and 25 root sections with 45 FLs were acquired. The width of FLs was from 3.43 µm to 143 µm; 32.2% of the points had width under 25 µm, while only 9.6% of them had width over 75 µm. The accuracy was 0.41, 0.54, 0.41, and 0.30 for NewTom VGi, Planmeca Promax 3D Max, Kavo 3D eXam, and Soredex Scanora3D, respectively. Conclusions: The modified temperature cycling method is a simple and effective method to establish narrow root fracture models, and the diagnosis efficiency for these narrow fracture lines was quite poor using all the four different CBCT units.

Using deep learning to distinguish malignant from benign parotid tumors on plain computed tomography images.

Objectives: Evaluating the diagnostic efficiency of deep-learning models to distinguish malignant from benign parotid tumors on plain computed tomography (CT) images. Materials and methods: The CT images of 283 patients with parotid tumors were enrolled and analyzed retrospectively. Of them, 150 were benign and 133 were malignant according to pathology results. A total of 917 regions of interest of parotid tumors were cropped (456 benign and 461 malignant). Three deep-learning networks (ResNet50, VGG16_bn, and DenseNet169) were used for diagnosis (approximately 3:1 for training and testing). The diagnostic efficiencies (accuracy, sensitivity, specificity, and area under the curve [AUC]) of three networks were calculated and compared based on the 917 images. To simulate the process of human diagnosis, a voting model was developed at the end of the networks and the 283 tumors were classified as benign or malignant. Meanwhile, 917 tumor images were classified by two radiologists (A and B) and original CT images were classified by radiologist B. The diagnostic efficiencies of the three deep-learning network models (after voting) and the two radiologists were calculated. Results: For the 917 CT images, ResNet50 presented high accuracy and sensitivity for diagnosing malignant parotid tumors; the accuracy, sensitivity, specificity, and AUC were 90.8%, 91.3%, 90.4%, and 0.96, respectively. For the 283 tumors, the accuracy, sensitivity, and specificity of ResNet50 (after voting) were 92.3%, 93.5% and 91.2%, respectively. Conclusion: ResNet50 presented high sensitivity in distinguishing malignant from benign parotid tumors on plain CT images; this made it a promising auxiliary diagnostic method to screen malignant parotid tumors.

Exploring the use of enhanced cone-beam CT technique to diagnose vertical root fracture.

OBJECTIVES: To explore the feasibility of using sodium iodide (NaI)+dimethyl sulfoxide (DMSO)+ethyl alcohol+ethyl acetate as a cone-beam CT (CBCT) contrast agent in the diagnosis of vertical root fracture (VRF). METHODS: 21 endodontically treated VRF teeth of 21 patients were collected in this study. All these 21 teeth were confirmed subtle fracture lines under transillumination, the number and position of fracture lines were recorded. All these patients had CBCT routine scanning (RS1) before extraction. After extraction, the teeth was performed micro-CT scanning and 3 in vitro CBCT scanning: CBCT routine scanning in vitro(RS2), CBCT enhanced scanning using meglumine diatrizoate (MD) as contrast agent(ES1); and CBCT enhanced scanning using NaI+DMSO+ethyl alcohol+ethyl acetate as contrast agent(ES2). The number of fracture lines was evaluated on all the 5 scanning modes and the accuracy of diagnosis was calculated. RESULTS: In all, there were 43 fracture lines on the 21 teeth. The accuracy of detection of fracture lines of CBCT RS1, RS2, ES1, ES2 and micro-CT was 0%, 20.9% (9/43), 11.6% (5/43), 93% (40/43) and 95.3% (41/43) respectively. Significant differences were found between ES2 vs. RS2, ES2 vs. ES1 (p < 0.01); however, no significant difference was found between ES2 vs. micro-CT (p > 0.05). CONCLUSION: CBCT enhanced scanning using NaI+DMSO+ethyl alcohol+ethyl acetate as contrast agent could be a prospective technique in the diagnosis of VRF.

Comparison of diagnosis of cracked tooth using contrast-enhanced CBCT and micro-CT.

OBJECTIVES: To evaluate the diagnostic accuracy using sodium iodide (NaI) and dimethyl sulfoxide (DMSO) as contrast agent in cone beam computed tomography (CBCT) scanning, and compare this with micro-CT. METHODS: 18 teeth were cracked artificially by soaking them cyclically in liquid nitrogen and hot water. After pre-treatment with artificial saliva, the teeth were scanned in four modes: CBCT routine scanning without contrast agent (RS); CBCT with meglumine diatrizoate (MD) as contrast agent (ES1); CBCT with NaI + DMSO as contrast agent (ES2); and micro-CT (mCT). The number of crack lines was evaluated in all four modes. Depth of crack lines and number of cracks presented from the occlusal surface to the pulp cavity (Np) in ES2 and micro-CT images were evaluated. RESULTS: There were 63 crack lines in all 18 teeth. 45 crack lines were visible on ES2 images as against four on the RS and ES1 images (p<0.05) and 37 on micro-CT images (p>0.05). Further, 34 crack lines could be observed on both ES2 and micro-CT images, and the average depth presented on ES2 images was 4.56 ± 0.88 mm and 3.89 ± 1.08 mm on micro-CT images (p<0.05). More crack lines could be detected from the occlusal surface to the pulp cavity on ES2 images than on micro-CT images (22 vs 11). CONCLUSION: CBCT with NaI +DMSO as the contrast agent was equivalent to micro-CT for number of crack lines and better for depth of crack lines. NaI + DMSO could be a potential CBCT contrast agent to improve diagnostic accuracy for cracked tooth.

Diagnosis of cracked teeth using cone-beam computed tomography: literature review and clinical experience.

Cone-beam computed tomography (CBCT) has been widely used in diagnosis of vertical root fractures (VRFs) in recent years. According to the American Association of Endodontists (AAE) classification, there are five types of cracked teeth and VRF is one of them. Due to the variability and overlapping of the cracks and fractures, some narrow fractures on the roots of VRFs could not be detected by CBCT, and some wide cracks on the crown of cracked teeth could be detected by CBCT. In this review, we firstly discussed the value of CBCT in the diagnosis of the AAE five types of cracked teeth and presented CBCT manifestations of some typical cases. Secondly, we summarized the factors influencing the diagnosis of cracks/fractures using CBCT, namely, CBCT device-related factors, patient-related factors, and evaluator-related factors. The possible strategies to improve the diagnostic accuracy in the clinic practice are also discussed in this part. Finally, we compared the differences of root fractures with lateral canals and external root resorption on CBCT images.