Limitation of the diagnostic value of MR images for diagnosing temporomandibular joint disorders

Automated pediatric TMJ articular disk identification and displacement classification in MRI with machine learning

OBJECTIVE

To evaluate the performance of an automated two-step model interpreting pediatric temporomandibular joint (TMJ) magnetic resonance imaging (MRI) using artificial intelligence (AI). Using deep learning techniques, the model first automatically identifies the disk and the TMJ osseous structures, and then an automated algorithm classifies disk displacement.

MATERIALS AND METHODS

MRI images of the TMJ from 235 pediatric patients (470 joints) were reviewed. TMJ structures were segmented, and the disk position was classified as dislocated or not dislocated. The UNet++ model was trained on MRI images from 135 and tested on images from 100 patients. Disk displacement was then classified by an automated algorithm assessing the location of disk centroid and surfaces for bone landmarks.

RESULTS

The mean age was 14.6 ± 0.1 years (Female: 138/235, 58 %), with 104 of 470 disks (22 %) anteriorly dislocated. UNet++ performed well in segmenting the TMJ anatomical structures, with a Dice coefficient of 0.67 for the disk, 0.91 for the condyle, and a Hausdorff distance of 2.8 mm for the articular eminence. The classification algorithm showed disk displacement classification comparable to human experts, with an AUC of 0.89-0.92 for the distance between the disk center and the eminence-condyle line.

CONCLUSION

A two-step automated model can accurately identify TMJ osseous structures and classify disk dislocation in pediatric TMJ MRI. This tool could assist clinicians who are not MRI experts when assessing pediatric TMJ disorders.

CLINICAL SIGNIFICANCE

Automated software that assists in locating the articular disk and surrounding structures and classifies disk displacement would improve the TMJ-MRI interpretation and the assessment of TMJ disorders in children.

Detecting the articular disk in magnetic resonance images of the temporomandibular joint using YOLO series

The purpose of this study was to construct an artificial intelligence object detection model to detect the articular disk from temporomandibular joint (TMJ) magnetic resonance (MR) images using YOLO series. The study included two experiments using datasets from different MR imaging machines. A total of 536 MR images were retrospectively examined. The performance of YOLOv5 and YOLOv8 in detecting the TMJ articular disk in both normal and displaced conditions was evaluated. The impact of image-processing techniques, such as histogram equalization (HE) and contrast-limited adaptive HE (CLAHE) on model performance, was also examined. The results showed that the YOLO series could detect the articular disk regardless of displacement, with superior performance on images of normal disk position. The results suggest the applicability of object detection models in improving the diagnosis of TMJ disorders.

Temporomandibular joint assessment in MRI images using artificial intelligence tools: where are we now? A systematic review

OBJECTIVES

To summarize the current evidence on the performance of artificial intelligence (AI) algorithms for the temporomandibular joint (TMJ) disc assessment and TMJ internal derangement diagnosis in magnetic resonance imaging (MRI) images.

METHODS

Studies were gathered by searching 5 electronic databases and partial grey literature up to May 27, 2024. Studies in humans using AI algorithms to detect or diagnose internal derangements in MRI images were included. The methodological quality of the studies was evaluated using the Quality Assessment Tool for Diagnostic of Accuracy Studies-2 (QUADAS-2) and a proposed checklist for dental AI studies.

RESULTS

Thirteen studies were included in this systematic review. Most of the studies assessed disc position. One study assessed disc perforation. A high heterogeneity related to the patient selection domain was found between the studies. The studies used a variety of AI approaches and performance metrics with CNN-based models being the most used. A high performance of AI models compared to humans was reported with accuracy ranging from 70% to 99%.

CONCLUSIONS

The integration of AI, particularly deep learning, in TMJ MRI, shows promising results as a diagnostic-assistance tool to segment TMJ structures and classify disc position. Further studies exploring more diverse and multicentre data will improve the validity and generalizability of the models before being implemented in clinical practice.

Evaluation of temporomandibular joint disc displacement with MRI-based radiomics analysis

OBJECTIVES

The purpose of this study was to propose a machine learning model and assess its ability to classify temporomandibular joint (TMJ) disc displacements on MR T1-weighted and proton density-weighted images.

METHODS

This retrospective cohort study included 180 TMJs from 90 patients with TMJ signs and symptoms. A radiomics platform was used to extract imaging features of disc displacements. Thereafter, different machine learning algorithms and logistic regression were implemented on radiomics features for feature selection, classification, and prediction. The radiomics features included first-order statistics, size- and shape-based features, and texture features. Six classifiers, including logistic regression, random forest, decision tree, k-nearest neighbours (KNN), XGBoost, and support vector machine were used for a model building which could predict the TMJ disc displacements. The performance of models was evaluated by sensitivity, specificity, and ROC curve.

RESULTS

KNN classifier was found to be the most optimal machine learning model for prediction of TMJ disc displacements. The AUC, sensitivity, and specificity for the training set were 0.944, 0.771, 0.918 for normal, anterior disc displacement with reduction (ADDwR) and anterior disc displacement without reduction (ADDwoR) while testing set were 0.913, 0.716, and 1 for normal, ADDwR, and ADDwoR. For TMJ disc displacements, skewness, root mean squared, kurtosis, minimum, large area low grey level emphasis, grey level non-uniformity, and long-run high grey level emphasis, were selected as optimal features.

CONCLUSIONS

This study has proposed a machine learning model by KNN analysis on TMJ MR images, which can be used for TMJ disc displacements.

Image preprocessing with contrast-limited adaptive histogram equalization improves the segmentation performance of deep learning for the articular disk of the temporomandibular joint on magnetic resonance images

OBJECTIVES

The objective was to evaluate the robustness of deep learning (DL)-based encoder-decoder convolutional neural networks (ED-CNNs) for segmenting temporomandibular joint (TMJ) articular disks using data sets acquired from 2 different 3.0-T magnetic resonance imaging (MRI) scanners using original images and images subjected to contrast-limited adaptive histogram equalization (CLAHE).

STUDY DESIGN

In total, 536 MR images from 49 individuals were examined. An expert orthodontist identified and manually segmented the disks in all images, which were then reviewed by another expert orthodontist and 2 expert oral and maxillofacial radiologists. These images were used to evaluate a DL-based semantic segmentation approach using an ED-CNN. Original and preprocessed CLAHE images were used to train and validate the models whose performances were compared.

RESULTS

Original and CLAHE images acquired on 1 scanner had pixel values that were significantly darker and with lower contrast. The values of 3 metrics-the Dice similarity coefficient, sensitivity, and positive predictive value-were low when the original MR images were used for model training and validation. However, these metrics significantly improved when images were preprocessed with CLAHE.

CONCLUSIONS

The robustness of the ED-CNN model trained on a dataset obtained from a single device is low but can be improved with CLAHE preprocessing. The proposed system provides promising results for a DL-based, fully automated segmentation method for TMJ articular disks on MRI.

Association between clinical symptoms and MRI image findings in symptomatic temporomandibular joint (TMJ) disease: A systematic review

To evaluate the association between clinical signs and symptoms of temporomandibular joint (TMJ) and magnetic resonance image (MRI) findings in patients with temporomandibular disorders (TMD). Relevant articles on humans over 18 years of age were obtained from five databases (Ovid MEDLINE, PubMed, Scopus, Web of Science, and Google Scholar) up to August 2022. Risk of bias assessment was completed using the Joanna Briggs Institute critical appraisal tools. The GRADEpro (Grading of Recommendations, Assessment, Development, and Evaluation) instrument was applied to assess the level of evidence across studies in a GRADE Summary of Findings table. In total, 22 studies were included in this systematic review. Of these, 11 studies highlighted that joint pain was positively associated with particular MRI findings: joint effusion, bone marrow edema, disk displacement with/without reduction, and condylar erosion. Masticatory muscle pain was found to have a strong positive correlation with disk displacement in four studies. Five studies found no significant association between MRI findings and masticatory muscle pain. Range of motion and MRI findings were examined in six studies. Limited mouth opening was found to be correlated with disk displacement in five studies. Of the 11 studies evaluating the correlation between joint noise and MRI findings, eight reported a significant association between disk displacement and TMJ noise. The results suggested that patients with joint pain and limited range of motion may benefit from MRI. Patients exhibiting primarily muscle pain are unlikely to benefit clinically from MRI. Future studies with improved quality are warranted.

Diagnostic accuracy of MRI-CBCT fused images in assessment of clinically diagnosed internal derangement of the temporomandibular joint

OBJECTIVES

To evaluate the diagnostic accuracy of Fused (MRI)-CBCT images in the assessment of internal derangement of the temporomandibular joint.

METHODS

MRI and CBCT images of the TMJ were evaluated bilaterally in 10 patients with clinically diagnosed internal derangement. Image fusion was performed using Amira 3D Software (version 5.4.3, Thermo Fisher Scientific Inc.).

RESULTS

The AUC index for MRI-CBCT fused images was 0.83, which was significantly different from the null hypothesis value of 0.5. This was confirmed by inter-examiner reliability index of 0.87, which is statistically significant.

CONCLUSION

MRI-CBCT fused images can significantly improve the accuracy and inter-examiner reliability in the evaluation of clinically diagnosed cases with internal derangement.

MRRead-A Novel Approach to Training Residents in Magnetic Resonance Image Temporomandibular Joint Interpretation

PURPOSE

The purpose of our study was to create an online, web-based training module that would instruct a group of participants in the interpretation of a magnetic resonance image (MRI) of the temporomandibular joint (TMJ) scan in a logical, step-wise manner to locate and identify all relevant features of internal derangement. The investigator's hypothesis was that implementing the MRRead TMJ training module would improve the participants' competency in the interpretation of MRI TMJ scans.

METHODS

The investigators designed and implemented a single-group prospective cohort study. The study population was composed of oral and maxillofacial surgery interns, residents, and staff. Subjects eligible for study inclusion were oral and maxillofacial surgeons of any level, between 18 and 50 years of age, that completed the MRRead training module to completion. The primary outcome variable was the difference between the pretest and post-test scores of the participants, and the frequency of missing internal derangement findings before and after the course. Secondary outcomes of interest were subjective data gathered from the course, including participant feedback as well as subjective evaluation of the training module and perceived benefit, as well as the learner's self-reported confidence level in interpreting MRI TMJ scans on their own before and after completion of the course. Descriptive and bivariate statistics were used.

RESULTS

The study sample consisted of 68 subjects, aged 20 to 47 (M = 29.1) years. When comparing the results of the exams precourse and post course, the overall frequency of missed features of internal derangement decreased from 19.7 to 5.9, and the total score overall increased from 8.5 to 68.6%. Regarding secondary outcomes, the majority of participants indicated that they agree or strongly agree with a number of positive subjective questions asked. In addition, there was a statistically significant increase in the participants' comfort levels in the interpretation of MRI TMJ scans.

CONCLUSION

The results of this study confirm the hypothesis that completing the MRRead training module (www.MRRead.ca) improves competency and comfort among participants in the interpretation of MRI TMJ scans and their identification of features of internal derangement correctly.

Introducing a novice-friendly classification system for magnetic resonance imaging of the temporomandibular joint disc morphology

OBJECTIVE

To introduce a new classification, aiming to correspond TMJ disc configuration with diagnosis, meanwhile reduce difficulty and subjectivity in TMJ MRI evaluation and training of TMD diagnosis for dental students.

METHODS

90 patients sought for TMD treatment were enrolled in the study, whose MRIs were used to establish the new classification. A total of 180 discs were evaluated using MRI for position (normal, DDWR or DDWoR) and classified by morphology according to previous (Murakami's classification) and new classification respectively. 60 discs were selected and judged by two groups (2 TMJ specialists and 30 dental students) to assess the reliability and validity of the new classification. Questionnaires were acquired for all observers to assess the attitude toward two classification systems. Descriptive statistics, Spearman's rank correlation coefficient, and intraclass correlation coefficient were performed. P < 0.05 was considered statistically significant.

RESULTS

In the new classification, Class 1 disc was significantly correlated with DDWR and Class 3 disc was significantly correlated with DDWoR. Interobserver reliability/consistency for observers between TMJ specialists was 0.867 when Murakami's classification was applied and 0.948 when the new classification was applied. Interobserver ICC value for dental students was 0.656 when using Murakami's classification, and 0.831 when using the new classification. The difference in attitude toward different classification systems was statistically significant.

CONCLUSION

A new classification of TMJ disc configuration is presented. The correlation between disc morphology and position revealed helps diagnosis and management. The new classification improves TMJ MRI interpreting accuracy and provides a better learning and using experience.

Temporomandibular joint segmentation in MRI images using deep learning

OBJECTIVES

Temporomandibular joint (TMJ) internal derangements (ID) represent the most prevalent temporomandibular joint disorder (TMD) in the population and its diagnosis typically relies on magnetic resonance imaging (MRI). TMJ articular discs in MRIs usually suffer from low resolution and contrast, and it is difficult to identify them. In this study, we applied two convolutional neural networks (CNN) to delineate mandibular condyle, articular eminence, and TMJ disc in MRI images.

METHODS

The models were trained on MRI images from 100 patients and validated on images from 40 patients using 2D slices and 3D volume as input, respectively. Data augmentation and five-fold cross-validation scheme were applied to further regularize the models. The accuracy of the models was then compared with four raters having different expertise in reading TMJ-MRI images to evaluate the performance of the models.

RESULTS

Both models performed well in segmenting the three anatomical structures. A Dice coefficient of about 0.7 for the articular disc, more than 0.9 for the mandibular condyle, and Hausdorff distance of about 2mm for the articular eminence were achieved in both models. The models reached near-expert performance for the segmentation of TMJ articular disc and performed close to the expert in the segmentation of mandibular condyle and articular eminence. They also surpassed non-experts in segmenting the three anatomical structures.

CONCLUSION

This study demonstrated that CNN-based segmentation models can be a reliable tool to assist clinicians identifying key anatomy on TMJ-MRIs. The approach also paves the way for automatic diagnosis of TMD.

CLINICAL SIGNIFICANCE

Accurately locating the articular disc is the hardest and most crucial step in the interpretation of TMJ-MRIs and consequently in the diagnosis of TMJ-ID. Automated software that assists in locating the articular disc and its surrounding structures would improve the reliability of TMJ-MRI interpretation, save time and assist in reader training. It will also serve as a foundation for additional automated analysis of pathology in TMJ structures to aid in TMD diagnosis.

Diagnostic efficacy of CBCT, MRI and CBCT-MRI fused images in determining anterior disc displacement and bone changes of temporomandibular joint

OBJECTIVES

To evaluate the diagnostic efficacy of CBCT-MRI fused image for anterior disc displacement and bone changes of temporomandibular joint (TMJ), which are the main imaging manifestations of temporomandibular disorders (TMD).

METHODS

Two hundred and thirty-one TMJs of 120 patients who were diagnosed with TMD were selected for the study. The anterior disc displacement, bone defect and bone hyperplasia evaluated by three experts were used as a reference standard. Three residents individually evaluated all the three sets of images, which were CBCT images, MRI images and CBCT-MRI fused images from individual CBCT and MRI images in a random order for the above-mentioned three imaging manifestations with a five-point scale. Each set of images was observed at least 1 week apart. A second evaluation was performed 4 weeks later. Intra- and interobserver agreements were assessed using the intraclass correlation coefficient (ICC). The areas under the ROC curves (AUCs) of the three image sets were compared with a Z test, and p < 0.05 was considered statistically significant.

RESULTS

One hundred and forty-five cases were determined as anterior disc displacement, 84 cases as bone defect and 40 cases as bone hyperplasia. The intra- and interobserver agreements in the CBCT-MRI fused image set (0.76-0.91) were good to excellent, and the diagnostic accuracy for bone changes was significantly higher than that of MRI image set (p＜0.05).

CONCLUSIONS

CBCT-MRI fused images can display the disc and surrounding bone structures simultaneously and significantly improve the observers' reliability and diagnostic accuracy, especially for inexperienced residents.

Automated segmentation of articular disc of the temporomandibular joint on magnetic resonance images using deep learning

Temporomandibular disorders are typically accompanied by a number of clinical manifestations that involve pain and dysfunction of the masticatory muscles and temporomandibular joint. The most important subgroup of articular abnormalities in patients with temporomandibular disorders includes patients with different forms of articular disc displacement and deformation. Here, we propose a fully automated articular disc detection and segmentation system to support the diagnosis of temporomandibular disorder on magnetic resonance imaging. This system uses deep learning-based semantic segmentation approaches. The study included a total of 217 magnetic resonance images from 10 patients with anterior displacement of the articular disc and 10 healthy control subjects with normal articular discs. These images were used to evaluate three deep learning-based semantic segmentation approaches: our proposed convolutional neural network encoder-decoder named 3DiscNet (Detection for Displaced articular DISC using convolutional neural NETwork), U-Net, and SegNet-Basic. Of the three algorithms, 3DiscNet and SegNet-Basic showed comparably good metrics (Dice coefficient, sensitivity, and positive predictive value). This study provides a proof-of-concept for a fully automated deep learning-based segmentation methodology for articular discs on magnetic resonance images, and obtained promising initial results, indicating that the method could potentially be used in clinical practice for the assessment of temporomandibular disorders.

Diagnostic correlation between clinical protocols and magnetic resonance findings in temporomandibular disorders: A systematic review and meta-analysis

BACKGROUND

Our aim was to assess the diagnostic correlation between clinical protocols and magnetic resonance (MRI) findings in temporomandibular disorders (TMDs), including disc displacement with and without reduction (DDwR; DDwoR) and arthralgia.

METHODS

A systematic review performed in two phases according to the PRISMA checklist. Specific indexing terms were used for search of studies assessing TMDs through clinical diagnostic protocols with the aid of Research Diagnostic Criteria for TMDs or Diagnostic Criteria for TMDs. Quality assessment performed using QUADAS-2. Heterogeneity was assessed using I² . Publication bias was assessed using funnel plots. For meta-analysis, we used random effect model or fixed effect. The main outcomes were sensitivity and specificity of clinical protocols.

RESULTS

Fourteen studies included in the qualitative analysis and 11 studies in the meta-analysis. None of the studies fulfilled all criteria of QUADAS-2. High heterogeneity and high publication bias were found among the studies. Clinical protocols for assessing DDwR compared with MRI showed pooled sensitivity of 66% and specificity of 72%. For DDwoR, sensitivity was 61% and specificity 98%. For arthralgia, sensitivity was 43% and specificity 68% for the presence of effusion.

CONCLUSIONS

This review reveals the need for studies with improved quality. Clinical protocols show poor to moderate validity in diagnosis of DDwR and DDwoR compared with MRI. No correlation was found between a clinical diagnosis of arthralgia and MRI effusion. Clinical diagnostic protocols can be used as screening tools, reserving the use of MRI for a more accurate diagnosis in patients with symptoms or dysfunction.

Diagnostic efficacy of CBCT, MRI, and CBCT-MRI fused images in distinguishing articular disc calcification from loose body of temporomandibular joint

OBJECTIVES

To evaluate the diagnostic efficacy of CBCT-MRI fused images for articular disc calcification of temporomandibular joint (TMJ).

MATERIALS AND METHODS

Twenty patients (24 TMJs) whose image examinations showed dense bodies in the TMJ space were included in the study. The locations of dense bodies evaluated by the three experts were used as a reference standard. Three oral and maxillofacial radiology residents evaluated whether the dense bodies were disc calcification or not, with a five-point scale for four sets of images (CBCT alone, MRI alone, both CBCT and MRI observed at a time, and CBCT-MRI fused images) randomly and independently. Each set of images was observed at least 1 week apart. A second evaluation was performed after 4 weeks. Intraclass correlation coefficients were calculated to assess the intra- and inter-observer agreement. The areas under the ROC curves (AUCs) were compared between the four image sets using Z test.

RESULTS

Ten cases were determined as articular disc calcifications, and fourteen cases were recognized as loose bodies in the TMJ spaces. The average AUC index for the CBCT-MRI fused images was 0.95 and significantly higher than the other sets (p < 0.01). The intra- and inter-observer agreement in the CBCT-MRI fused images (0.90-0.91, 0.93) was excellent and higher than those in the other images.

CONCLUSIONS

CBCT-MRI fused images can significantly improve the observers' reliability and accuracy in determining articular disc calcification of the TMJ.

CLINICAL RELEVANCE

The multimodality image fusion is feasible in detecting articular disc calcification of the TMJ which are hard to define by CBCT or MRI alone. It can be utilized especially for inexperienced residents to shorten the learning curve and improve diagnostic accuracy.

Three-Dimensional Assessment of Temporomandibular Joint Using MRI-CBCT Image Registration

Purpose

To introduce a new approach to reconstruct a 3D model of the TMJ using magnetic resonance imaging (MRI) and cone-beam computed tomography (CBCT) registered images, and to evaluate the intra-examiner reproducibility values of reconstructing the 3D models of the TMJ.

Methods

MRI and CBCT images of five patients (10 TMJs) were obtained. Multiple MRIs and CBCT images were registered using a mutual information based algorithm. The articular disc, condylar head and glenoid fossa were segmented at two different occasions, at least one-week apart, by one investigator, and 3D models were reconstructed. Differences between the segmentation at two occasions were automatically measured using the surface contours (Average Perpendicular Distance) and the volume overlap (Dice Similarity Index) of the 3D models. Descriptive analysis of the changes at 2 occasions, including means and standard deviation (SD) were reported to describe the intra-examiner reproducibility.

Results

The automatic segmentation of the condyle revealed maximum distance change of 1.9±0.93 mm, similarity index of 98% and root mean squared distance of 0.1±0.08 mm, and the glenoid fossa revealed maximum distance change of 2±0.52 mm, similarity index of 96% and root mean squared distance of 0.2±0.04 mm. The manual segmentation of the articular disc revealed maximum distance change of 3.6±0.32 mm, similarity index of 80% and root mean squared distance of 0.3±0.1 mm.

Conclusion

The MRI-CBCT registration provides a reliable tool to reconstruct 3D models of the TMJ’s soft and hard tissues, allows quantification of the articular disc morphology and position changes with associated differences of the condylar head and glenoid fossa, and facilitates measuring tissue changes over time.

Usefulness of MRI-CBCT image registration in the evaluation of temporomandibular joint internal derangement by novice examiners

OBJECTIVE

To assess whether novice examiners can more reliably determine temporomandibular joint (TMJ) disk derangements with the use of fused images of magnetic resonance imaging and cone beam computed tomography (MRI-CBCT) compared with MRI images alone.

STUDY DESIGN

Thirty dental students with minimal exposure to TMJ imaging received a 30-minute calibration session in which TMJ diagnostic imaging and the normal anatomy of the TMJ internal derangement were explained. The students evaluated the disk positions of 16 TMJs in two sets of images (MRI images alone and MRI-CBCT images) randomly and independently. The disk positions evaluated by two experienced radiologists were used for comparison.

RESULTS

The internal consistency among all students improved from an unacceptable consistency (α = 0.40) with MRI images alone to a good consistency (α = 0.84) with MRI-CBCT images. The agreement in evaluation between the students and the radiologists improved from a poor agreement with MRI images alone (k mean = 0.07 ± 0.12) to a moderate agreement with MRI-CBCT images (k mean = 0.55 ± 0.25). This improvement in the agreement was significant (P < .001).

CONCLUSIONS

The use of fused MRI and CBCT images to visualize the TMJ in a single display significantly improved the examiners' reliability and accuracy of assessment of disk positions. The improvement of novice readers in assessing the disk positions highlights the potential use of MRI-CBCT image fusion as an educational tool.

MRI alone versus MRI-CBCT registered images to evaluate temporomandibular joint internal derangement

OBJECTIVE

To evaluate the effect of magnetic resonance imaging-cone beam computed tomography (MRI-CBCT) image registration on inter- and intraexaminer consistency when evaluating temporomandibular joint (TMJ) internal derangement compared to MRI alone.

METHODS

MRI and CBCT images of 25 patients (50 TMJs) were obtained and coregistered using mutual-information rigid image registration via Mirada XD software. Two experienced radiologists independently and blindly evaluated two types of images (MRI alone and MRI-CBCT registered images) at two different times (T1 and T2) for TMJ internal derangement, based on sagittal and coronal articular disc position in relation to the head of the condyle and the posterior slope of the articular eminence.

RESULTS

The intraexaminer consistency with MRI alone (examiner 1 = 0.85 [0.74-0.92]; examiner 2 = 0.91 [0.84-0.95]) was lower than for the MRI-CBCT registered images (examiner 1 = 0.95 [0.91-0.97]; examiner 2 = 0.97 [0.96-0.99]). The interexaminer consistency of evaluating internal derangement with MRI alone (0.52 [0.18-0.73] at T1; 0.71 [0.45-0.84] at T2) was lower than for the MRI-CBCT registered images (0.97 [0.95-0.98] at T1; 0.98 [0.96-0.99] at T2). When disc position classification was dichotomized to normal versus anteriorly displaced, intraexaminer agreement for the two examiners was 0.52 and 0.63 for MRI alone, but was 0.91 and 0.92 for MRI-CBCT registered images. Interexaminer agreement for MRI alone was 0.29 at T1 and 0.42 at T2, but was 0.96 at both examination times for MRI-CBCT registered images.

CONCLUSION

The MRI-CBCT registered images improved intra- and interexaminer consistency in the evaluation of internal derangement of TMJ.

Magnetic resonance imaging of temporomandibular joint: morphometric study of asymptomatic volunteers

BACKGROUND

The aims of this study were to determine the best suited magnetic resonance imaging scanning plane, scanning sequence, and imaging modality for the evaluation of the temporomandibular joint (TMJ) and quantitatively assess the relationship of articular disk position to condyle position.

METHODS

One hundred four TMJs in 52 symptom-free heads were examined by magnetic resonance imaging. The best scanning plane, scanning sequence, and scanning parameter were determined according to the imaging time and image quality. Bilateral symmetry of the articular disk and mandibular condyle was measured by using the automatic measurement of 3.0-T GE Excite Signa MR scanner.

RESULTS

Fast spin-echo sequence, oblique sagittal imaging plane, and proton density imaging were the best suited scanning sequence, scanning planes, and imaging modality, respectively. The thicknesses of the anterior and posterior bands and for the intermediate zone were not statistically different for both sides. The posterior band of the disk was found to originate in an area adjacent to the 12-o'clock position of the condyle (± 5 degrees), whereas the anterior band of the disk originated adjacent to 1-o'clock position (28 ± 6 degrees). The anteroposterior diameter and mediolateral diameter of the condylar processes were not statistically different for both sides. The axial condylar angle between the plane of the greatest mediolateral diameter of the condylar processes and the midsagittal plane were also not statistically different for both sides.

CONCLUSIONS

The magnetic resonance images can depict clearly major regional anatomic structures and position in the TMJ, which can be used in the early diagnosis for the TMJ disorder.

Stomach position in prediction of survival in left-sided congenital diaphragmatic hernia with or without fetoscopic endoluminal tracheal occlusion

OBJECTIVE

To investigate the value of fetal stomach position in predicting postnatal outcome in left-sided congenital diaphragmatic hernia (CDH) with and without fetoscopic endoluminal tracheal occlusion (FETO).

METHODS

This was a retrospective review of CDH cases that were expectantly managed or treated with FETO, assessed from May 2008 to October 2013, in which we graded, on a scale of 1-4, stomach position on the four-chamber view of the heart with respect to thoracic structures. Logistic regression analysis was used to investigate the effect of management center (Paris, Brussels, Barcelona, Milan), stomach grading, observed-to-expected lung area-to-head circumference ratio (O/E-LHR), gestational age at delivery, birth weight in expectantly managed CDH, gestational ages at FETO and at removal and period of tracheal occlusion, on postnatal survival in CDH cases treated with FETO.

RESULTS

We identified 67 expectantly managed CDH cases and 47 CDH cases that were treated with FETO. In expectantly managed CDH, stomach position and O/E-LHR predicted postnatal survival independently. In CDH treated with FETO, stomach position and gestational age at delivery predicted postnatal survival independently.

CONCLUSION

In left-sided CDH with or without FETO, stomach position is predictive of postnatal survival.

Bayesian belief network analysis applied to determine the progression of temporomandibular disorders using MRI

OBJECTIVES

This study investigated the applicability of a Bayesian belief network (BBN) to MR images to diagnose temporomandibular disorders (TMDs). Our aim was to determine the progression of TMDs, focusing on how each finding affects the other.

METHODS

We selected 1.5-T MRI findings (33 variables) and diagnoses (bone changes and disc displacement) of patients with TMD from 2007 to 2008. There were a total of 295 cases with 590 sides of temporomandibular joints (TMJs). The data were modified according to the research diagnostic criteria of TMD. We compared the accuracy of the BBN using 11 algorithms (necessary path condition, path condition, greedy search-and-score with Bayesian information criterion, Chow-Liu tree, Rebane-Pearl poly tree, tree augmented naïve Bayes model, maximum log likelihood, Akaike information criterion, minimum description length, K2 and C4.5), a multiple regression analysis and an artificial neural network using resubstitution validation and 10-fold cross-validation.

RESULTS

There were 191 TMJs (32.4%) with bone changes and 340 (57.6%) with articular disc displacement. The BBN path condition algorithm using resubstitution validation and 10-fold cross-validation was >99% accurate. However, the main advantage of a BBN is that it can represent the causal relationships between different findings and assign conditional probabilities, which can then be used to interpret the progression of TMD.

CONCLUSIONS

Osteoarthritic bone changes progressed from condyle to articular fossa and finally to mandibular bone contours. Disc displacement was directly related to severe bone changes. Early bone changes were not directly related to disc displacement. TMJ functional factors (condylar translation, bony space and disc form) and age mediated between bone changes and disc displacement.

Contrast-enhanced microCT (EPIC-μCT) ex vivo applied to the mouse and human jaw joint

OBJECTIVES

The temporomandibular joint (TMJ) is susceptive to the development of osteoarthritis (OA). More detailed knowledge of its development is essential to improve our insight into TMJ-OA. It is imperative to have a standardized reliable three-dimensional (3D) imaging method that allows for detailed assessment of both bone and cartilage in healthy and diseased joints. We aimed to determine the applicability of a contrast-enhanced microCT (µCT) technique for ex vivo research of mouse and human TMJs.

METHODS

Equilibrium partitioning of an ionic contrast agent via µCT (EPIC-µCT) was previously applied for cartilage assessment in the knee joint. The method was ex vivo, applied to the mouse TMJ and adapted for the human TMJ.

RESULTS

EPIC-µCT (30-min immersion time) was applied to mouse mandibular condyles, and 3D imaging revealed an average cartilage thickness of 110 ± 16 µm. These measurements via EPIC-µCT were similar to the histomorphometric measures (113 ± 19 µm). For human healthy OA-affected TMJ samples, the protocol was adjusted to an immersion time of 1 h. 3D imaging revealed a significant thicker cartilage layer in joints with early signs of OA compared with healthy joints (414.2 ± 122.6 and 239.7 ± 50.5 µm, respectively). A subsequent significant thinner layer was found in human joints with late signs of OA (197.4 ± 159.7 µm).

CONCLUSIONS

The EPIC-µCT technique is effective for the ex vivo assessment of 3D cartilage morphology in the mouse as well as human TMJ and allows bone-cartilage interaction research in TMJ-OA.

Web-based calibration of observers using MRI of the temporomandibular joint

OBJECTIVES

Research diagnostic criteria for temporomandibular disorders (RDC/TMDs) were proposed in 1992 with the aim of standardizing and testing methods for diagnosing TMDs. RDC/TMDs have so far been lacking standardized methods for imaging and criteria for imaging diagnosis of disorders of the temporomandibular joint (TMJ). Criteria for disorders of the TMJ have recently been proposed for MRI. The aim of this study was to publish MR images of the TMJ on the web and to test the agreement of five observers using the criteria.

METHODS

20 cases of MRI of the TMJ were published on the web together with the criteria. The MR images were taken in closed and open mouth positions, and sagittal and coronal views. Five observers diagnosed disc position, disc shape, joint effusion and loose calcified bodies in the TMJ.

RESULTS

In all cases except one, three or more observers agreed upon the diagnosis. All agreed on whether a loose calcified body was present or not. The second best agreement was obtained for disc position in the sagittal view, where all observers agreed in 16 of the 20 cases. For disc position in the coronal view and the evaluation of the disc shape, observer agreement was lower.

CONCLUSION

Criteria were useful in order to standardize and simplify evaluation and thereby probably increase the diagnostic outcome among different observers for MRI of the TMJ. We recommend that the criteria be used internationally to facilitate comparisons between different studies.

Magnetic resonance imaging in the normal fetal heart and in congenital heart disease

OBJECTIVE

To evaluate prospectively the feasibility of magnetic resonance imaging (MRI) for assessment of the fetal heart for congenital heart disease (CHD).

METHODS

This was a cross-sectional study, including 66 fetuses with a normal heart and 40 with CHD. The fetal heart was examined on MRI using axial steady-state free precession (SSFP) sequences. Regression analysis was used to investigate the effect on the ability to visualize cardiac anatomy of gestational age at examination, maternal body mass index, presence of fetal cardiac abnormality, fetal movements, fetal lie and twinning. The sensitivity and specificity of detecting cardiac defects were calculated.

RESULTS

The four-chamber view was visualized in 98.1% of fetuses. The sensitivity of detecting a cardiac defect on the four-chamber view was 88% and the specificity 96%. The ability to visualize the left and right outflow tracts was only influenced by the presence of fetal movements: for the left outflow tract 94.4 vs. 50.0% visualization and for the right outflow tract 92.6 vs. 53.8% visualization without and with fetal movements, respectively. The sensitivity of detecting a cardiac defect of the left outflow tract was 63% and the specificity 100%, while sensitivity and specificity were 59 and 97%, respectively, for the right outflow tract.

CONCLUSIONS

Despite the use of SSFP sequences, MRI in the fetal heart remains of limited value. It can only be used as a second-line approach for abnormalities of the four-chamber view suspected at prenatal ultrasound.

Evaluation of the reproducibility in the interpretation of magnetic resonance images of the temporomandibular joint

OBJECTIVES

The aim was to determine the intra- and interexaminer reproducibility in the interpretation of MRI of the temporomandibular joint among independent observers, with respect to six specific articular characteristics, and to discover which of these had greater and lesser agreement.

METHODS

30 magnetic resonance examinations of temporomandibular joints of adults were independently interpreted by 9 experienced and trained observers at 2 different times. Observers were divided into three groups according to their specialties: surgeon dentists specialized in temporomandibular dysfunction and orofacial pain, surgeon dentists specialized in radiology and medical doctors specialized in radiology. The reproducibility analysis was carried out using Cohen's kappa coefficient.

RESULTS

The interexaminer reproducibility ranged from slight to fair. The intraexaminer reproducibility ranged from slight to no agreement. In the interexaminer evaluation, anterior disc displacement without reduction presented greater agreement, whereas change in condylar head shape showed the poorest agreement. In the intraexaminer evaluation, anterior disc displacement without reduction presented slight agreement, whereas, for the other characteristics, no agreement was observed.

CONCLUSION

Examiners do not demonstrate reproducibility in the interpretation of MRI of temporomandibular joints. Therefore, more efforts are necessary with respect to understanding the changes that may be detected in these images in terms of diagnosis and appropriate treatment approaches.

Grid analysis improves reliability in follicle counts made by ultrasonography in women with polycystic ovary syndrome

Poor reliability has been reported when counting the total number of follicles in polycystic ovaries using conventional two-dimensional (2-D) ultrasound viewing methods. In the current study, we report good reliability in follicle counts when observers imposed a programmable grid system over the viewing window. Four observers estimated total follicle counts in 45 ovarian ultrasound scans by compartmentalizing the ovary into 9 to 12 grid sections and performing focused follicle counts per section. The mean number of follicles counted per ovary was 44.6 +/- 2.3. The level of inter-observer agreement when making follicle counts was 0.82 and total follicle counts did not differ among observers. The level of intra-observer agreement was 0.93 which further corroborated the utility of this method for making dependable follicle counts. In summary, the ability to obtain reproducible follicle counts will help to establish reliable diagnostic criteria for polycystic ovarian morphology.

Assessment of ultrasonographic features of polycystic ovaries is associated with modest levels of inter-observer agreement

Background

There is growing acceptance that polycystic ovaries are an important marker of polycystic ovary syndrome (PCOS) despite significant variability when making the ultrasound diagnosis. To better understand the nature of this variability, we proposed to evaluate the level of inter-observer agreement when identifying and quantifying individual ultrasonographic features of polycystic ovaries.

Methods

Digital recordings of transvaginal ultrasound scans performed in thirty women with PCOS were assessed by four observers with training in Radiology or Reproductive Endocrinology. Observers evaluated the scans for: 1) number of follicles ≥ 2 mm per ovary, 2) largest follicle diameter, 3) ovarian volume, 4) follicle distribution pattern and 5) presence of a corpus luteum (CL). Lin's concordance correlation coefficients and kappa statistics for multiple raters were used to assess inter-observer agreement.

Results

Agreement between observers ranged from 0.08 to 0.63 for follicle counts, 0.27 to 0.88 for largest follicle diameter, 0.63 to 0.86 for ovarian volume, 0.51 to 0.76 for follicle distribution pattern and 0.76 to 0.90 for presence of a CL. Overall, reproductive endocrinologists demonstrated better agreement when evaluating ultrasonographic features of polycystic ovaries compared to radiologists (0.71 versus 0.53; p = 0.04).

Conclusion

Inter-observer agreement for assessing ultrasonographic features of polycystic ovaries was moderate to poor. These findings support the need for standardized training modules to characterize polycystic ovarian morphology on ultrasonography.

The frequency and extent of mammillary body atrophy associated with surgical removal of a colloid cyst

BACKGROUND AND PURPOSE

Patients who have had a colloid cyst removed from the third ventricle sometimes experience some difficulty with day-to-day memory. This study provided quantitative MR imaging volume measures of 1 structure potentially responsible for mnemonic problems, the mammillary bodies. Additional volume estimates in structures connected to the mammillary bodies sought to determine the specificity of any atrophy.

MATERIALS AND METHODS

Volume estimates of the mammillary bodies were performed on 38 patients after surgical removal of colloid cysts and 20 control subjects by the application of stereologic volume-estimation techniques. For the mammillary body measures, 2 groups of MR images were assessed (0.8- and 1.0-mm section thickness) to compare the sensitivity of each imaging sequence for detecting any atrophy. Other structures associated with memory processes, such as the hippocampus and fornix, were also assessed quantitatively to determine whether there was a correlation between mammillary body damage and atrophy in connecting structures.

RESULTS

Our investigations established the superiority of 0.8-mm-volume scans over standard isotropic 1.0-mm-thick-volume scans for mammillary body assessments. Comparisons with 20 age-matched controls revealed that patients with colloid cysts frequently showed significant mammillary body atrophy (mean volume of colloid cysts, 0.037 cm(3) right and 0.038 cm(3) left; control subjects, 0.069 cm(3) right and 0.067 cm(3) left). In fact, every patient had a mammillary body volume below the control mean, and the majority of patients had a volume decrease of >1 SD (82% right, 74% left). Mammillary body volumes correlated with fornix volumes in the same patient group.

CONCLUSIONS

Our results reveal the frequent presence of mammillary body atrophy in patients with surgical removal of colloid cysts and indicate that this atrophy is partly due to a loss of temporal lobe projections in the fornix.

Improving inter-observer variability in the evaluation of ultrasonographic features of polycystic ovaries

BACKGROUND

We recently reported poor inter-observer agreement in identifying and quantifying individual ultrasonographic features of polycystic ovaries. Our objective was to determine the effect of a training workshop on reducing inter-observer variation in the ultrasonographic evaluation of polycystic ovaries.

METHODS

Transvaginal ultrasound recordings from thirty women with polycystic ovary syndrome (PCOS) were evaluated by three radiologists and three reproductive endocrinologists both before and after an ultrasound workshop. The following endpoints were assessed: 1) follicle number per ovary (FNPO), 2) follicle number per single cross-section (FNPS), 3) largest follicle diameter, 4) ovarian volume, 5) follicle distribution pattern and 6) presence of a corpus luteum (CL). Lin's concordance correlation coefficients (rho) and kappa statistics for multiple raters (kappa) were used to assess level of inter-observer agreement (>0.80 good, 0.60 - 0.80 moderate/fair, <0.60 poor).

RESULTS

Following the workshop, inter-observer agreement improved for the evaluation of FNPS (rho = 0.70, delta rho = +0.11), largest follicle diameter (rho = 0.77, delta rho = +0.10), ovarian volume (rho = 0.84, delta rho = +0.12), follicle distribution pattern (kappa = 0.80, delta kappa = +0.21) and presence of a CL (kappa = 0.87, delta kappa = +0.05). No improvement was evident for FNPO (rho = 0.54, delta rho = -0.01). Both radiologists and reproductive endocrinologists demonstrated improvement in scores (p < 0.001).

CONCLUSION

Reliability in evaluating ultrasonographic features of polycystic ovaries can be significantly improved following participation in a training workshop. If ultrasonographic evidence of polycystic ovaries is to be used as an objective measure in the diagnosis of PCOS, then standardized training modules should be implemented to unify the approach to evaluating polycystic ovarian morphology.