Radiograph-based rheumatoid arthritis diagnosis via convolutional neural network.

OBJECTIVES: Rheumatoid arthritis (RA) is a severe and common autoimmune disease. Conventional diagnostic methods are often subjective, error-prone, and repetitive works. There is an urgent need for a method to detect RA accurately. Therefore, this study aims to develop an automatic diagnostic system based on deep learning for recognizing and staging RA from radiographs to assist physicians in diagnosing RA quickly and accurately. METHODS: We develop a CNN-based fully automated RA diagnostic model, exploring five popular CNN architectures on two clinical applications. The model is trained on a radiograph dataset containing 240 hand radiographs, of which 39 are normal and 201 are RA with five stages. For evaluation, we use 104 hand radiographs, of which 13 are normal and 91 RA with five stages. RESULTS: The CNN model achieves good performance in RA diagnosis based on hand radiographs. For the RA recognition, all models achieve an AUC above 90% with a sensitivity over 98%. In particular, the AUC of the GoogLeNet-based model is 97.80%, and the sensitivity is 100.0%. For the RA staging, all models achieve over 77% AUC with a sensitivity over 80%. Specifically, the VGG16-based model achieves 83.36% AUC with 92.67% sensitivity. CONCLUSION: The presented GoogLeNet-based model and VGG16-based model have the best AUC and sensitivity for RA recognition and staging, respectively. The experimental results demonstrate the feasibility and applicability of CNN in radiograph-based RA diagnosis. Therefore, this model has important clinical significance, especially for resource-limited areas and inexperienced physicians.

Enchondroma Detection from Hand Radiographs with an Interactive Deep Learning Segmentation Tool-A Feasibility Study.

Enchondromas are common benign bone tumors, usually presenting in the hand. They can cause symptoms such as swelling and pain but often go un-noticed. If the tumor expands, it can diminish the bone cortices and predispose the bone to fracture. Diagnosis is based on clinical investigation and radiographic imaging. Despite their typical appearance on radiographs, they can primarily be misdiagnosed or go totally unrecognized in the acute trauma setting. Earlier applications of deep learning models to image classification and pattern recognition suggest that this technique may also be utilized in detecting enchondroma in hand radiographs. We trained a deep learning model with 414 enchondroma radiographs to detect enchondroma from hand radiographs. A separate test set of 131 radiographs (47% with an enchondroma) was used to assess the performance of the trained deep learning model. Enchondroma annotation by three clinical experts served as our ground truth in assessing the deep learning model's performance. Our deep learning model detected 56 enchondromas from the 62 enchondroma radiographs. The area under receiver operator curve was 0.95. The F1 score for area statistical overlapping was 69.5%. Our deep learning model may be a useful tool for radiograph screening and raising suspicion of enchondroma.

Hand and Wrist Involvement in Seropositive Rheumatoid Arthritis, Seronegative Rheumatoid Arthritis, and Psoriatic Arthritis-The Value of Classic Radiography.

The hand and wrist are among the most common anatomical areas involved in rheumatic diseases, especially seropositive and seronegative rheumatoid arthritis (RA) and psoriatic arthritis (PsA). The purpose of this study was to identify the most differentiating radiographic characteristics of PsA, seropositive RA, and seronegative RA, particularly in the early stages. A retrospective analysis of radiographic hand findings was performed on 180 seropositive RA patients (29 males, 151 females, mean age at the point of acquisition of the analyzed radiograph of 53.4 y/o, SD 12.6), 154 PsA patients (45 males, 109 females, age median of 48.1 y/o, SD 12.4), and 36 seronegative RA patients (4 males, 32 females, age median of 53.1 y/o, SD 17.1) acquired during the period 2005-2020. Posterior-anterior and Nørgaard views were analyzed in all patients. The radiographs were evaluated for three radiographic findings: type of symmetry (asymmetric/bilateral/changes in corresponding joint compartments/'mirror-image' symmetry), anatomic location (e.g., wrist, metacarpophalangeal (MCP), proximal interphalangeal (PIP), distal interphalangeal (DIP) joints), and type of lesions (e.g., juxta-articular osteoporosis, bone cysts, erosions, proliferative bone changes). The study showed that symmetric distribution of lesions defined as 'lesions present in corresponding compartments' was more suggestive of seropositive or seronegative RA than PsA. Lesions affecting the PIP joints, wrist, or styloid process of the radius; juxta-articular osteoporosis, joint space narrowing, joint subluxations, or dislocations were more common in patients with seropositive RA than in those with PsA, whereas DIP joints' involvement and proliferative bone changes were more likely to suggest PsA than seropositive RA. Lesions in PIP, MCP, and wrist joints, as well as erosions, advanced bone damage, joint subluxations, dislocations, and joint space narrowing, were more common in seropositive RA patients than in seronegative RA patients. The ulnar styloid was more commonly affected in seronegative RA patients than in PsA patients. The study confirmed that types of bone lesions and their distribution in the hands and wrists can be useful in differentiating seropositive RA from PsA and suggests that seronegative RA varies in radiological presentation from seropositive RA and PsA.

High performance for bone age estimation with an artificial intelligence solution.

PURPOSE: The purpose of this study was to compare the performance of an artificial intelligence (AI) solution to that of a senior general radiologist for bone age assessment. MATERIAL AND METHODS: Anteroposterior hand radiographs of eight boys and eight girls from each age interval between five and 17 year-old from four different radiology departments were retrospectively collected. Two board-certified pediatric radiologists with knowledge of the sex and chronological age of the patients independently estimated the Greulich and Pyle bone age to determine the standard of reference. A senior general radiologist not specialized in pediatric radiology (further referred to as "the reader") then determined the bone age with knowledge of the sex and chronological age. The results of the reader were then compared to those of the AI solution using mean absolute error (MAE) in age estimation. RESULTS: The study dataset included a total of 206 patients (102 boys of mean chronological age of 10.9 ± 3.7 [SD] years, 104 girls of mean chronological age of 11 ± 3.7 [SD] years). For both sexes, the AI algorithm showed a significantly lower MAE than the reader (P < 0.007). In boys, the MAE was 0.488 years (95% confidence interval [CI]: 0.28-0.44; r2 = 0.978) for the AI algorithm and 0.771 years (95% CI: 0.64-0.90; r2 = 0.94) for the reader. In girls, the MAE was 0.494 years (95% CI: 0.41-0.56; r2 = 0.973) for the AI algorithm and 0.673 years (95% CI: 0.54-0.81; r2 = 0.934) for the reader. CONCLUSION: The AI solution better estimates the Greulich and Pyle bone age than a general radiologist does.

Sex and External Size Specific Limitations in Assessing Bone Health From Adult Hand Radiographs.

Morphological parameters measured for the second metacarpal from hand radiographs are used clinically for assessing bone health during growth and aging. Understanding how these morphological parameters relate to metacarpal strength and strength at other anatomical sites is critical for providing informed decision-making regarding treatment strategies and effectiveness. The goals of this study were to evaluate the extent to which 11 morphological parameters, nine of which were measured from hand radiographs, relate to experimentally measured whole-bone strength assessed at multiple anatomical sites and to test whether these associations differed between men and women. Bone morphology and strength were assessed for the second and third metacarpals, radial diaphysis, femoral diaphysis, and proximal femur for 28 white male donors (18-89 years old) and 35 white female donors (36-89+ years old). The only morphological parameter to show a significant correlation with strength without a sex-specific effect was cortical area. Dimensionless morphological parameters derived from hand radiographs correlated significantly with strength for females, but few did for males. Males and females showed a significant association between the circularity of the metacarpal cross-section and the outer width measured in the mediolateral direction. This cross-sectional shape variation contributed to systematic bias in estimating strength using cortical area and assuming a circular cross-section. This was confirmed by the observation that use of elliptical formulas reduced the systematic bias associated with using circular approximations for morphology. Thus, cortical area was the best predictor of strength without a sex-specific difference in the correlation but was not without limitations owing to out-of-plane shape variations. The dependence of cross-sectional shape on the outer bone width measured from a hand radiograph may provide a way to further improve bone health assessments and informed decision making for optimizing strength-building and fracture-prevention treatment strategies. © 2022 The Authors. JBMR Plus published by Wiley Periodicals LLC on behalf of American Society for Bone and Mineral Research.

Development of a computer-aided quality assurance support system for identifying hand X-ray image direction using deep convolutional neural network.

The convenience of imaging has improved with digitization; however, there has been no progress in the methods used to prevent human error. Therefore, radiographic incidents and accidents are not prevented. In Japan, image interpretation is conducted for incident prevention; nevertheless, in some cases, incidents are overlooked. Thus, assistance from a computer-aided quality assurance support system is important. This study developed a method to identify hand image direction, which is an elementary technology of a computer-aided quality assurance support system. In total, 14,236 hand X-ray images were used to classify hand directions (upward, downward, rightward, and leftward) commonly evaluated in clinical settings. The accuracy of the conventional classification method using original images, classification method with histogram equation images, and a novel classification method using binarization images for background removal via U-Net segmentation was evaluated. The following classification accuracy rates were achieved: 89.20% if the original image was input, 99.10% if the histogram equation image was input, and 99.70% if binarization images for background removal via U-Net segmentation was input. Our computer-aided quality assurance support system can be used to identify hand direction with high accuracy.

Sequential spine-hand radiography for assessing skeletal maturity with low radiation EOS imaging system for bracing treatment recommendation in adolescent idiopathic scoliosis: a feasibility and validity study.

PURPOSE: The EOS-imaging system is increasingly adopted for clinical follow-up in scoliosis with the advantages of simultaneous biplanar imaging of the spine in an erect position. Skeletal maturity assessment using a hand radiograph is an essential adjunct to spinal radiography in scoliosis follow-up. This study aims at testing the feasibility and validity of a newly proposed EOS workflow with sequential spine-hand radiography for skeletal maturity assessment and bracing recommendation. METHODS: EOS spine-hand radiographs from patients with diagnosis of idiopathic scoliosis, including both sexes and an age range of ten to 14 years, were scored using the Thumb Ossification Composite Index (TOCI), Sanders and Risser methods. Intraclass correlation coefficients (ICCs) were calculated for inter/intraobserver agreement and were tested with Cronbach's alpha values. RESULTS: In all, 60 EOS-spine hand radiographs selected from subjects with diagnosis of adolescent idiopathic scoliosis (AIS), including 32 male patients (mean age 11.53 years; 10 to 14) and 28 female patients (mean age 11.50 years; 10 to 13) who underwent sequential spine-hand low dose EOS imaging were generated for analysis. The overall interobserver (ICC = 0.997) and intraobserver agreement (α > 0.9) demonstrated excellent agreement for TOCI staging; ICC > 0.994 for both TOCI and Sanders staging comparing traditional digital versus EOS hand radiography; ICC ≥ 0.841 for agreement on bracing recommendation among TOCI versus the Risser and Sanders system. CONCLUSION: With the proposed new EOS workflow it was feasible to produce high image quality for skeletal maturity assessment with excellent reliability and validity to inform consistent bracing recommendation in AIS. The workflow is applicable for busy daily clinic settings in tertiary scoliosis centres with reduced time cost, improved efficiency and throughput of the radiology department. LEVEL OF EVIDENCE: III.

Combining wrist age and third molars in forensic age estimation: how to calculate the joint age estimate and its error rate in age diagnostics.

BACKGROUND: Forensic age estimation in living adolescents is based on several methods, e.g. the assessment of skeletal and dental maturation. Combination of several methods is mandatory, since age estimates from a single method are too imprecise due to biological variability. The correlation of the errors of the methods being combined must be known to calculate the precision of combined age estimates. AIM: To examine the correlation of the errors of the hand and the third molar method and to demonstrate how to calculate the combined age estimate. SUBJECTS AND METHODS: Clinical routine radiographs of the hand and dental panoramic images of 383 patients (aged 7.8-19.1 years, 56% female) were assessed. RESULTS: Lack of correlation (r = -0.024, 95% CI = -0.124 to + 0.076, p = 0.64) allows calculating the combined age estimate as the weighted average of the estimates from hand bones and third molars. Combination improved the standard deviations of errors (hand = 0.97, teeth = 1.35 years) to 0.79 years. CONCLUSION: Uncorrelated errors of the age estimates obtained from both methods allow straightforward determination of the common estimate and its variance. This is also possible when reference data for the hand and the third molar method are established independently from each other, using different samples.

Age estimation based on bone length using 12 regression models of left hand X-ray images for Asian children below 19 years old.

Age estimation was used in forensic anthropology to help in the identification of individual remains and living person. However, the estimation methods tend to be unique and applicable only to a certain population. This paper analyzed age estimation using twelve regression models carried out on X-ray images of the left hand taken from an Asian data set for subjects under the age of 19. All the nineteen bones of the left hand were measured using free image software and the statistical analysis were performed using SPSS. There are two methods to determine age in this study which are single bone method and all bones method. For single bone method, S-curve regression model was found to have the highest R-square value using second metacarpal for males, and third proximal phalanx for females. For age estimation using single bone, fifth metacarpal from males and fifth proximal phalanx from females can be used due to the lowest mean square error (MSE) value. To conclude, multiple linear regressions is the best techniques for age estimation in cases where all bones are available, but if not, S-curve regression can be used using single bone method.