Deep learning-based super-resolution of structural brain MRI at 1.5 T: application to quantitative volume measurement

OBJECTIVE

This study investigated the feasibility of using deep learning-based super-resolution (DL-SR) technique on low-resolution (LR) images to generate high-resolution (HR) MR images with the aim of scan time reduction. The efficacy of DL-SR was also assessed through the application of brain volume measurement (BVM).

MATERIALS AND METHODS

In vivo brain images acquired with 3D-T1W from various MRI scanners were utilized. For model training, LR images were generated by downsampling the original 1 mm-2 mm isotropic resolution images. Pairs of LR and HR images were used for training 3D residual dense net (RDN). For model testing, actual scanned 2 mm isotropic resolution 3D-T1W images with one-minute scan time were used. Normalized root-mean-square error (NRMSE), peak signal-to-noise ratio (PSNR), and structural similarity (SSIM) were used for model evaluation. The evaluation also included brain volume measurement, with assessments of subcortical brain regions.

RESULTS

The results showed that DL-SR model improved the quality of LR images compared with cubic interpolation, as indicated by NRMSE (24.22% vs 30.13%), PSNR (26.19 vs 24.65), and SSIM (0.96 vs 0.95). For volumetric assessments, there were no significant differences between DL-SR and actual HR images (p > 0.05, Pearson's correlation > 0.90) at seven subcortical regions.

DISCUSSION

The combination of LR MRI and DL-SR enables addressing prolonged scan time in 3D MRI scans while providing sufficient image quality without affecting brain volume measurement.

Does Articular Disc Position Change Following Mandibular Setback Surgery?

BACKGROUND

Maintaining condyle position following bilateral sagittal split ramus osteotomy (BSSO) is crucial to minimizing postoperative relapse. However, the impact of BSSO on the articular disc position remains inconclusive.

PURPOSE

This study aimed to investigate the changes in articular disc position following setback BSSO surgery.

STUDY DESIGN, SETTING, AND SAMPLING

In this prospective cohort study, subjects with mandibular prognathism requiring setback BSSO were enrolled between August 2021 and June 2022 at the Oral and Maxillofacial Surgery Clinic, Faculty of Dentistry, Chiang Mai University, Thailand. Patients with surgical complications, loss of follow-up, or significant artifacts in their MR images were excluded.

PREDICTOR/EXPOSURE/INDEPENDENT VARIABLES

The predictor variable was time. The articular disc position was assessed at 3 time points, preoperatively (T0), 3 months postsurgery (T1), and 6 months postsurgery (T2).

MAIN OUTCOME

The primary and secondary outcome variables were the changes in articular disc position between T0-T2 and T0-T1, respectively. Articular disc position was coded as normal, anterior disc displacement with reduction (ADDwR), anterior disc displacement without reduction (ADDwoR), and anterior disc displacement without reduction and degenerative joint disease (ADDwoR + DJD).

COVARIATES

Covariate variables collected included age (years), sex (male or female), asymmetry (present or absent), surgical procedure (single jaw (BSSO) or bimaxillary surgery), and setback distance (millimeters).

ANALYSES

Friedman's test with 80% power and a significance level of 0.05 was employed. Pairwise comparisons were performed using the Dunn-Bonferroni posthoc test to identify statistically significant differences.

RESULTS

The study included 16 subjects, 6 females and 10 males, with a total of 32 TMJs. Subjects had a mean age of 23.75 (4.57) years. The proportion of TMJs with normal disc position postoperatively increased from 3 (9.4%) to 19 (59.4%). Statistically significant differences were found in the changes in disc position over time (P < .001).

CONCLUSIONS

Following setback BSSO, the articular discs underwent changes, with a majority of ADDwR cases transitioning to a normal position. Cases with ADDwoR also demonstrated disc reduction capability after surgery. The combined orthodontic treatment and setback BSSO appear to have an effect on articular disc position in skeleton class III patients.

Enhancing Caries Detection in Bitewing Radiographs Using YOLOv7

The study aimed to evaluate the impact of image size, area of detection (IoU) thresholds and confidence thresholds on the performance of the YOLO models in the detection of dental caries in bitewing radiographs. A total of 2575 bitewing radiographs were annotated with seven classes according to the ICCMS™ radiographic scoring system. YOLOv3 and YOLOv7 models were employed with different configurations, and their performances were evaluated based on precision, recall, F1-score and mean average precision (mAP). Results showed that YOLOv7 with 640 × 640 pixel images exhibited significantly superior performance compared to YOLOv3 in terms of precision (0.557 vs. 0.268), F1-score (0.555 vs. 0.375) and mAP (0.562 vs. 0.458), while the recall was significantly lower (0.552 vs. 0.697). The following experiment found that the overall mAPs did not significantly differ between 640 × 640 pixel and 1280 × 1280 pixel images, for YOLOv7 with an IoU of 50% and a confidence threshold of 0.001 (p = 0.866). The last experiment revealed that the precision significantly increased from 0.570 to 0.593 for YOLOv7 with an IoU of 75% and a confidence threshold of 0.5, but the mean-recall significantly decreased and led to lower mAPs in both IoUs. In conclusion, YOLOv7 outperformed YOLOv3 in caries detection and increasing the image size did not enhance the model's performance. Elevating the IoU from 50% to 75% and confidence threshold from 0.001 to 0.5 led to a reduction of the model's performance, while simultaneously improving precision and reducing recall (minimizing false positives and negatives) for carious lesion detection in bitewing radiographs.

A web application for sex and stature estimation from radiographic proximal femur for a Thai population

In both forensic and archaeological domains, the discovery of incomplete human remains is a frequent occurrence. Nevertheless, the estimation of biological profiles from such remains presents a challenge due to the absence of crucial skeletal elements, such as the skull and pelvis. This study aimed to assess the utility of the proximal femur in the forensic identification process by creating a web application for osteometric analysis of the proximal femur. The aim was to determine the sex and stature of an individual from radiographs of the left anteroposterior femur. To accomplish this, an automated method was developed for acquiring linear measurements from radiographic images of the proximal femur using Python tools. The application of Hough techniques and Canny edge detection was utilized to generate linear femoral dimensions from radiographs. A total of 354 left femora were radiographed and measured by the algorithm. The sex classification model employed in this study was the Naïve Bayes algorithm (accuracy = 91.2 %). Results indicated that Gaussian process regression (GPR) was the most effective method for estimating stature (mean error = 4.68 cm, SD = 3.93 cm). The proposed web application holds the potential to serve as a valuable asset in the realm of forensic investigations in Thailand, particularly in the estimation of biological profiles from fragmentary skeletal remains.

Effects of manual therapy combined with therapeutic exercise versus routine physical therapy on brain biomarkers in patients with chronic non-specific neck pain in Thailand: a study protocol for a single-blinded randomised controlled trial

INTRODUCTION

Structural brain alterations in pain-related areas have been demonstrated in patients with non-specific neck pain. While manual therapy combined with therapeutic exercise is an effective management for neck pain, its underlying mechanisms are poorly understood. The primary objective of this trial is to investigate the effects of manual therapy combined with therapeutic exercise on grey matter volume and thickness in patients with chronic non-specific neck pain. The secondary objectives are to assess changes in white matter integrity, neurochemical biomarkers, clinical features of neck pain, cervical range of motion and cervical muscle strength.

METHODS AND ANALYSIS

This study is a single-blinded, randomised controlled trial. Fifty-two participants with chronic non-specific neck pain will be recruited into the study. Participants will be randomly allocated to either an intervention or control group (1:1 ratio). Participants in the intervention group will receive manual therapy combined with therapeutic exercise for 10 weeks (two visits per week). The control group will receive routine physical therapy. Primary outcomes are whole-brain and regional grey matter volume and thickness. Secondary outcomes are white matter integrity (fractional anisotropy and mean diffusivity), neurochemical biomarkers (N-acetylaspartate, creatine, glutamate/glutamine, myoinositol and choline), clinical features (neck pain intensity, duration, neck disability and psychological symptoms), cervical range of motion and cervical muscle strength. All outcome measures will be taken at baseline and postintervention.

ETHICS AND DISSEMINATION

Ethical approval of this study has been granted by Faculty of Associated Medical Science, Chiang Mai University. The results of this trial will be disseminated through a peer-reviewed publication.

TRIAL REGISTRATION NUMBER

NCT05568394.

Variability of gait characteristics in lameness elephant

Lameness has a significant impact not only on the economy but also on elephant welfare. Several gait characteristics are altered to compensate for the discomfort. The traditional approach to detecting lameness has relied on mahout and veterinarian observation. The study aimed to determine how lameness affected the variability of an elephant's gait by using a three-dimensional inertial measurement unit (IMU) with Wi-Fi sensors. Five elephants with lameness, as determined by an experienced veterinarian and two, non-lamed elephants, aged between 58-80 years were included in the study. Gait biomechanics including limb segment motion, obtained from individually gyrometric- and accelero-based parameters and demonstrated as a graphical pattern showing changes in absolute rotation angle over time. The result revealed some character changes in gait kinematics parameters, but it was heterogeneity with an inconclusive pattern. The interlimb coordination could be a part of maintaining the actual locomotion pattern, or it could be a result of the mild degree of lameness for which all of the clients have compensated. This study introduces a new objective method for quantifying gait changes caused by lameness, additional research is required to make this objective more clinically applicable.

Feasibility of deep learning for dental caries classification in bitewing radiographs based on the ICCMS™ radiographic scoring system

OBJECTIVE

To evaluate the potential of deep learning models for categorization of dental caries in bitewing radiographs based on the International Caries Classification and Management System (ICCMS™) radiographic scoring system (RSS).

STUDY DESIGN

In total, 2758 annotated bitewing radiographs were randomly divided into 3 experiments to assess the ResNet-18, -50, -101, and -152. Experiment A tested 4-class ICCMS™-RSS training and validation using Carestream (CS) radiographs; experiment B tested training and validation using CS and VistaScan radiographs; experiment C tested 7-class ICCMS™-RSS training and validation using CS and VistaScan radiographs. The performance matrices and the areas under the receiver operating characteristic curves were analyzed to assess all procedures.

RESULTS

In experiment A, ResNet-50 and ResNet-152 were equally accurate (71.11%) and approximately 78% sensitive. The latter presented the highest specificity (56.90%). In experiment B, ResNet-50 presented the highest sensitivity (79.51%) but ResNet-152 had the highest specificity (60.71%). In experiment C, all models markedly underperformed in distinguishing the 7-class ICCMS™-RSS with specificities of 16.46% to 22.41%. They had fewer classification errors in the 4-class classification (28.89%-35.56%) than in the 7-class classification (42.34%-53.06%). The areas under the receiver operating characteristic curves of all models were unanimously comparable.

CONCLUSIONS

The ResNet models were able to classify dental caries according to the ICCMS™-RSS with average performances. The models underperformed in complicated classification tasks.

Assessment of YOLOv3 for caries detection in bitewing radiographs based on the ICCMS™ radiographic scoring system

OBJECTIVES

To assess the feasibility of the YOLOv3 model under the intersection over union (IoU) thresholds of 0.5 (IoU₅₀) and 0.75 (IoU₇₅) for caries detection in bitewing radiographs based on the International Caries Classification and Management System (ICCMS™).

MATERIALS AND METHODS

We trained the YOLOv3 model by feeding 994 annotated radiographs with the IoU₅₀ and IoU₇₅ thresholds. The testing procedure (n = 175) was subsequently conducted to evaluate the model's prediction metrics on caries classification based on the ICCMS™ radiographic scoring system.

RESULTS

Regarding the 4-class classification representing caries severity, YOLOv3 could accurately detect and classify enamel caries and initial dentin caries (class RA) (IoU₅₀ vs IoU₇₅: precision, 0.75 vs 0.71; recall, 0.67 vs 0.64). Concerning the 7-class classification signifying specific caries depth (class 0, healthy tooth; classes RA1-3, initial caries affecting outer half, inner half of enamel, and the outer 1/3 of dentin; class RB4, caries extending to the middle 1/3 of dentin; classes RC5-6, extensively cavitated caries affecting the inner 1/3 of dentin and involving the pulp chamber), YOLOv3 could accurately detect and classify caries with pulpal exposure (class RC6) (IoU₅₀ vs IoU₇₅: precision, 0.77 vs 0.73; recall, 0.61 vs 0.57) but it failed to predict the outer half of enamel caries (class RA1) (IoU₅₀ vs IoU₇₅: precision, 0.35 vs 0.32; recall, 0.23 vs 0.21).

CONCLUSIONS

YOLOv3 yielded acceptable performances in both IoU₅₀ and IoU₇₅. Although the performance metrics decreased in the 7-class detection, the two thresholds revealed comparable results. However, the model could not consistently detect initial-stage caries affecting the outermost surface of the enamel.

CLINICAL RELEVANCE

YOLOv3 could be implemented to detect and classify dental caries according to the ICCMS™ classification with acceptable performances to assist dentists in making treatment decisions.

Fully automated method for dental age estimation using the ACF detector and deep learning

Background

Dental age estimation plays an important role in identifying an unknown person. In forensic science, estimating age with high accuracy depends on the experience of the practitioner. Previous studies proposed classification of tooth development of the mandibular third molar by following Demirjian’s method, which is useful for dental age estimation. Although stage of tooth growth is very helpful in assessing age estimation, it must be performed manually. The drawback of this procedure is its need for skilled observers to carry out the tasks precisely and reproducibly because it is quite detailed. Therefore, this research aimed to apply computer-aid methods for reducing time and subjectivity in dental age estimation by using dental panoramic images based on Demirjian’s method. Dental panoramic images were collected from persons aged 15 to 23 years old. In accordance with Demirjian’s method, this study focused only on stages D to H of tooth development, which were discovered in the 15- to 23-year age range. The aggregate channel features detector was applied automatically to localize and crop only the lower left mandibular third molar in panoramic images. Then, the convolutional neural network model was applied to classify cropped images into D to H stages. Finally, the classified stages were used to estimate dental age.

Results

Experimental results showed that the proposed method in this study can localize the lower left mandibular third molar automatically with 99.5% accuracy, and training in the convolutional neural network model can achieve 83.25% classification accuracy using the transfer learning strategy with the Resnet50 network.

Conclusion

In this work, the aggregate channel features detector and convolutional neural network model were applied to localize a specific tooth in a panoramic image and identify the developmental stages automatically in order to estimate the age of the subjects. The proposed method can be applied in clinical practice as a tool that helps clinicians to reduce the time and subjectivity for dental age estimation.

An automated liver segmentation in liver iron concentration map using fuzzy c-means clustering combined with anatomical landmark data

Background

To estimate median liver iron concentration (LIC) calculated from magnetic resonance imaging, excluded vessels of the liver parenchyma region were defined manually. Previous works proposed the automated method for excluding vessels from the liver region. However, only user-defined liver region remained a manual process. Therefore, this work aimed to develop an automated liver region segmentation technique to automate the whole process of median LIC calculation.

Methods

553 MR examinations from 471 thalassemia major patients were used in this study. LIC maps (in mg/g dry weight) were calculated and used as the input of segmentation procedures. Anatomical landmark data were detected and used to restrict ROI. After that, the liver region was segmented using fuzzy c-means clustering and reduced segmentation errors by morphological processes. According to the clinical application, erosion with a suitable size of the structuring element was applied to reduce the segmented liver region to avoid uncertainty around the edge of the liver. The segmentation results were evaluated by comparing with manual segmentation performed by a board-certified radiologist.

Results

The proposed method was able to produce a good grade output in approximately 81% of all data. Approximately 11% of all data required an easy modification step. The rest of the output, approximately 8%, was an unsuccessful grade and required manual intervention by a user. For the evaluation matrices, percent dice similarity coefficient (%DSC) was in the range 86–92, percent Jaccard index (%JC) was 78–86, and Hausdorff distance (H) was 14–28 mm, respectively. In this study, percent false positive (%FP) and percent false negative (%FN) were applied to evaluate under- and over-segmentation that other evaluation matrices could not handle. The average of operation times could be reduced from 10 s per case using traditional method, to 1.5 s per case using our proposed method.

Conclusion

The experimental results showed that the proposed method provided an effective automated liver segmentation technique, which can be applied clinically for automated median LIC calculation in thalassemia major patients.

Impact of Weight Carriage on Joint Kinematics in Asian Elephants Used for Riding

Simple Summary

Riding elephants is one of the most controversial activities in the tourist industry, with concerns over whether load carrying is physically harmful. Here, we used an empirical approach to test how carrying loads up to 15% of the elephant’s body mass affected gait parameters. The maximal angles of fore- and hindlimb joints of elephants walking at a normal, self-selected speed carrying a mahout only were first evaluated and then compared to those walking with a saddle carrying two people plus added weight to reach a 15% body mass load. Data were analyzed using a computerized three-dimensional inertial measurement system. There were no significant differences between movement angles, including flexion, extension, abduction, and adduction of the fore- or hindlimbs between these two riding conditions. Thus, we found no evidence that carrying two people in a saddle causes significant changes in gait patterns or potentially affects musculoskeletal function. More studies are needed to further test longer durations of riding on different types of terrain to develop appropriate working guidelines for captive elephants. Nevertheless, elephants appear capable of carrying significant amounts of weight on the back without showing signs of physical distress.

Abstract

Background: Elephants in Thailand have changed their roles from working in the logging industry to tourism over the past two decades. In 2020, there were approximately 2700 captive elephants participating in activities such as riding and trekking. During work hours, riding elephants carry one or two people in a saddle on the back with a mahout on the neck several hours a day and over varying terrain. A concern is that this form of riding can cause serious injuries to the musculoskeletal system, although to date there have been no empirical studies to determine the influence of weight carriage on kinematics in elephants. Methods: Eight Asian elephants from a camp in Chiang Mai Province, Thailand, aged between 21 and 41 years with a mean body mass of 3265 ± 140.2 kg, were evaluated under two conditions: walking at a normal speed without a saddle and with a 15% body mass load (saddle and two persons plus additional weights). Gait kinematics, including the maximal angles of fore- and hindlimb joints, were determined using a novel three-dimensional inertial measurement system with wireless sensors. Results: There were no statistical differences between movement angles and a range of motion of the fore- and hindlimbs, when an additional 15% of body mass was added. Conclusion: There is no evidence that carrying a 15% body mass load causes significant changes in elephant gait patterns. Thus, carrying two people in a saddle may have minimal effects on musculoskeletal function. More studies are needed to further test longer durations of riding on different types of terrain to develop appropriate working guidelines for captive elephants. Nevertheless, elephants appear capable of carrying significant amounts of weight on the back without showing signs of physical distress.

Feasibility of accelerated 3D T1-weighted MRI using compressed sensing: application to quantitative volume measurements of human brain structures

OBJECTIVE

Scan time reduction is necessary for volumetric acquisitions to improve workflow productivity and to reduce motion artifacts during MRI procedures. We explored the possibility that Compressed Sensing-4 (CS-4) can be employed with 3D-turbo-field-echo T1-weighted (3D-TFE-T1W) sequence without compromising subcortical measurements on clinical 1.5 T MRI.

MATERIALS AND METHODS

Thirty-three healthy volunteers (24 females, 9 males) underwent imaging scans on a 1.5 T MRI equipped with a 12-channel head coil. 3D-TFE-T1W for whole-brain coverage was performed with different acceleration factors, including SENSE-2, SENSE-4, CS-4. Freesurfer, FSL's FIRST, and volBrain packages were utilized for subcortical segmentation. All processed data were assessed using the Wilcoxon signed-rank test.

RESULTS

The results obtained from SENSE-2 were considered as references. For SENSE-4, the maximum signal-to-noise ratio (SNR) drop was detected in the Accumbens (51.96%). For CS-4, the maximum SNR drop was detected in the Amygdala (10.55%). Since the SNR drop in CS-4 is relatively small, the SNR in all of the subcortical volumes obtained from SENSE-2 and CS-4 are not statistically different (P > 0.05), and their Pearson's correlation coefficients are larger than 0.90. The maximum biases of SENSE-4 and CS-4 were found in the Thalamus with the mean of differences of 1.60 ml and 0.18 ml, respectively.

CONCLUSION

CS-4 provided sufficient quality of 3D-TFE-T1W images for 1.5 T MRI equipped with a 12-channel receiver coil. Subcortical volumes obtained from the CS-4 images are consistent among different post-processing packages.

Automatic cardiac T2* relaxation time estimation from magnetic resonance images using region growing method with automatically initialized seed points

BACKGROUND AND OBJECTIVE

Heart failure due to iron-overload cardiomyopathy is one of the main causes of mortality. The cardiomyopathy is reversible if intensive iron chelation treatment is done in time, but the diagnosis is often delayed because the cardiac iron deposition is unpredictable and the symptoms are lately detected. There are many ways to assess iron-overload. However, the widely used and approved method is by using MRI which is performed by calculating the T2* (T2-star). In order to compute the T2* value, the region of interest (ROI) is manually selected by an expert which may require considerable time and skills. The aim of this work is hence to develop the cardiac T2* measurement by using region growing algorithm for automatically segmenting the ROI in cardiac MR images. Mathematical morphologies are also used to reduce some errors.

METHODS

Thirty MR images with free-breathing and respiratory-trigger technique were used in this work. The segmentation algorithm yields good results when compared with the manual segmentation performed by two experts.

RESULTS

The averages of positive predictive value, the sensitivity, the Hausdorff distance, and the Dice similarity coefficient are 0.76, 0.84, 7.78 pixels, and 0.80 when compared with the two experts' opinions. The T2* values were carried out based on the automatically segmented ROI's. The mean difference of T2* values between the proposed technique and the experts' opinion is about 1.40ms.

CONCLUSIONS

The results demonstrate the accuracy of the proposed method in T2* value estimation. Some previous methods were implemented for comparisons. The results show that the proposed method yields better segmentation and T2* value estimation performances.