Search

Objective: To examine the genetic architecture of cam morphology using alpha angle (AA) as a proxy measure and conduct an AA genome-wide association study (GWAS) followed by Mendelian randomization (MR) to evaluate its causal relationship with hip osteoarthritis (OA). Methods: Observational analyses examined associations between AA measurements derived from hip dual x-ray absorptiometry (DXA) scans from the UK Biobank study and radiographic hip OA outcomes and subsequent total hip replacement. Following these analyses, an AA GWAS meta-analysis was performed (N = 44,214) using AA measurements previously derived in the Rotterdam Study. Linkage disequilibrium score regression assessed the genetic correlation between AA and hip OA. Genetic associations considered significant (P < 5 × 10−8) were used as AA genetic instrument for 2-sample MR analysis. Results: DXA-derived AA showed expected associations between AA and radiographic hip OA (adjusted odds ratio [OR] 1.63 [95% confidence interval (95% CI) 1.58, 1.67]) and between AA and total hip replacement (adjusted hazard ratio 1.45 [95% CI 1.33, 1.59]) in the UK Biobank study cohort. The heritability of AA was 10%, and AA had a moderate genetic correlation with hip OA (rg = 0.26 [95% CI 0.10, 0.43]). Eight independent genetic signals were associated with AA. Two-sample MR provided weak evidence of causal effects of AA on hip OA risk (inverse variance weighted OR 1.84 [95% CI 1.14, 2.96], P = 0.01). In contrast, genetic predisposition for hip OA had stronger evidence of a causal effect on increased AA (inverse variance weighted β = 0.09 [95% CI 0.04, 0.13], P = 4.58 × 10−5). Conclusion: Expected observational associations between AA and related clinical outcomes provided face validity for the DXA-derived AA measurements. Evidence of bidirectional associations between AA and hip OA, particularly for risk of hip OA on AA, suggests that hip shape modeling secondary to a genetic predisposition to hip OA contributes to the well-established relationship between hip OA and cam morphology in older adults.

ObjectiveMinimum joint space width (mJSW) from 2-dimensional images provides a proxy for cartilage thickness. This study aimed to conduct a genome-wide association study (GWAS) of mJSW to (i) identify new genetic determinants of mJSW and use them to (ii) examine causal effects of mJSW on hip osteoarthritis (HOA) risk.MethodsGWAS meta-analysis of hip mJSW derived from plain X-rays (four cohorts) or DXA (one cohort) was performed, stratified by sex and adjusted for age and ancestry principal components. Mendelian randomisation (MR) and cluster analyses were used to examine causal effect of mJSW on HOA.Results50,745 individuals were included in the meta-analysis. 42 SNPs, which mapped to 39 loci (35 novel), were identified. Mendelian randomisation (MR) revealed little evidence of a causal effect of mJSW on HOA (βIVW-0.01 [95% CI -0.19, 0.17]). However, MR-Clust analysis suggested the null MR estimates reflected the net effect of two distinct causal mechanisms cancelling each other out, one of which was protective, whereas the other increased HOA susceptibility. For the latter mechanism, all loci were positively associated with height, suggesting mechanisms leading to greater height and mJSW increase the risk of HOA in later life.ConclusionsGWAS and MR analyses suggested one group of mJSW loci reduces HOA risk via increased mJSW, suggesting possible utility as targets for chondroprotective therapies. The second group of mJSW loci increased HOA risk, despite increasing mJSW, but were also positively related to height, suggesting they contribute to mJSW and HOA risk via a growth-related mechanism.

Objectives To examine whether acetabular dysplasia (AD), cam and/or pincer morphology are associated with radiographic hip osteoarthritis (rHOA) and hip pain in UK Biobank (UKB) and, if so, what distribution of osteophytes is observed.DesignParticipants from UKB with a left hip dual-energy x-ray absorptiometry (DXA) scan had alpha angle (AA), lateral centre-edge angle (LCEA) and joint space narrowing (JSN) derived automatically. Cam and pincer morphology, and AD were defined using AA and LCEA. Osteophytes were measured manually and rHOA grades were calculated from JSN and osteophyte measures. Logistic regression was used to examine the relationships between these hip morphologies and rHOA, osteophytes, JSN, and hip pain.Results6,807 individuals were selected (mean age: 62.7; 3382/3425 males/females). Cam morphology was more prevalent in males than females (15.4% and 1.8% respectively). In males, cam morphology was associated with rHOA [OR 3.20 (95% CI 2.41-4.25)], JSN [1.53 (1.24-1.88)], and acetabular [1.87 (1.48-2.36)], superior [1.94 (1.45-2.57)] and inferior [4.75 (3.44-6.57)] femoral osteophytes, and hip pain [1.48 (1.05-2.09)]. Broadly similar associations were seen in females, but with weaker statistical evidence. Neither pincer morphology nor AD showed any associations with rHOA or hip pain.ConclusionsCam morphology was predominantly seen in males in whom it was associated with rHOA and hip pain. In males and females, cam morphology was associated with inferior femoral head osteophytes more strongly than those at the superior femoral head and acetabulum. Further studies are justified to characterise the biomechanical disturbances associated with cam morphology, underlying the observed osteophyte distribution.

Background/Aims Radiographic hip osteoarthritis (rHOA) is traditionally defined on hip x-rays, using subjective methods such as Kellgren-Lawrence scoring. Associations between subjective rHOA measures and symptoms are inconsistent. Applying digital tools to high-resolution dual-energy X-ray absorptiometry (DXA) scans, we aimed to develop a novel semi-automated classifier for rHOA and evaluate the face validity of the classifier based on relationships with hip pain, hospital diagnosed OA (HES OA), and risk of total hip replacement (THR). Methods Using hip DXAs in UK Biobank, osteophyte grades 0-3 were assigned based on manually measured osteophyte area. Minimum joint space width (mJSW) was automatically measured using outline points placed by a machine learning-based algorithm and then used to categorise individuals into joint space narrowing (JSN) grades 0-3. Osteophyte and JSN grades were combined, using a novel system giving greater to weight to osteophytes, to categorise individuals into rHOA grades 0-4. Logistic regression giving odds ratios (OR) was used to examine associations between rHOA grade and hip pain, and HES OA. Cox proportional hazard models giving hazard ratios (HR) were used to examine associations between rHOA grade and subsequent THR. Our adjusted model included age, sex, height and weight as covariates. Results 40,340 individuals were included in the study (mean age 63.7 [range 44-82], 19294/21046 male/female). 32758 (81.2%) had rHOA grade 0, 4565 (11.3%) grade 1, 2317 (5.7%) grade 2, 543 (1.3%) grade 3, 157 (0.4%) grade 4, with all features of rHOA being more common in males than females. rHOA grades ≥2 were associated with all three clinical outcomes in both unadjusted and adjusted models, a clear dose-response relationship was seen with each increase in grade showing a large rise in OR/HRs (Table 1). Grade 4 rHOA was strongly predictive of THR (HR 57.70 [95%CI 38.08-87.44]). Conclusion We successfully applied a novel semi-automated classifier to over 40,000 individuals from UKB. The validity of our classifier was supported by the strong and progressive relationships observed between rHOA and hip pain, and HES OA and risk of THR. We conclude that hip DXAs provide a promising means of defining rHOA, with potential screening applications in the clinic. Disclosure B.G. Faber: None. R. Ebsim: None. F.R. Saunders: None. M. Frysz: None. C. Lindner: None. J.S. Gregory: None. R.M. Aspden: None. N.C. Harvey: None. G. Davey Smith: None. T. Cootes: None. J.H. Tobias: None.

Objective Conventional scoring methods for radiographic hip OA (rHOA) are subjective and show inconsistent relationships with clinical outcomes. To provide a more objective rHOA scoring method, we aimed to develop a semi-automated classifier based on DXA images and confirm its relationships with clinical outcomes. Methods Hip DXAs in UK Biobank (UKB) were marked up for osteophyte area from which acetabular, superior and inferior femoral head osteophyte grades were derived. Joint space narrowing (JSN) grade was obtained automatically from minimum joint space width (mJSW) measures. Clinical outcomes related to rHOA comprised hip pain, hospital diagnosed OA (HES OA) and total hip replacement. Logistic regression and Cox proportional hazard modelling were used to examine associations between overall rHOA grade (0–4; derived from combining osteophyte and JSN grades) and the clinical outcomes. Results A toal of 40 340 individuals were included in the study (mean age 63.7), of whom 81.2% had no evidence of rHOA, while 18.8% had grade ≥1 rHOA. Grade ≥1 osteophytes at each location and JSN were associated with hip pain, HES OA and total hip replacement. Associations with all three clinical outcomes increased progressively according to rHOA grade, with grade 4 rHOA and total hip replacement showing the strongest association [57.70 (38.08–87.44)]. Conclusions Our novel semi-automated tool provides a useful means for classifying rHOA on hip DXAs, given its strong and progressive relationships with clinical outcomes. These findings suggest DXA scanning can be used to classify rHOA in large DXA-based cohort studies supporting further research, with the future potential for population-based screening.

Introduction: Alpha angle (AA) is a widely used imaging measure of hip shape that is commonly used to define cam morphology, a bulging of the lateral aspect of the femoral head. Cam morphology has shown strong associations with hip osteoarthritis (OA) making the AA a clinically relevant measure. In both clinical practice and research studies, AA tends to be measured manually which can be inconsistent and time-consuming. Objective: We aimed to (i) develop an automated method of deriving AA from anterior-posterior dual-energy x-ray absorptiometry (DXA) scans; and (ii) validate this method against manual measures of AA. Methods: 6,807 individuals with left hip DXAs were selected from UK Biobank. Outline points were manually placed around the femoral head on 1,930 images before training a Random Forest-based algorithm to place the points on a further 4,877 images. An automatic method for calculating AA was written in Python 3 utilising these outline points. An iterative approach was taken to developing and validating the method, testing the automated measures against independent batches of manually measured images in sequential experiments. Results: Over the course of six experimental stages the concordance correlation coefficient, when comparing the automatic AA to manual measures of AA, improved from 0.28 [95% confidence interval 0.13-0.43] for the initial version to 0.88 [0.84-0.92] for the final version. The inter-rater kappa statistic comparing automatic versus manual measures of cam morphology, defined as AA ³≥60°, improved from 0.43 [80% agreement] for the initial version to 0.86 [94% agreement] for the final version. Conclusions: We have developed and validated an automated measure of AA from DXA scans, showing high agreement with manually measuring AA. The proposed method is available to the wider research community from Zenodo.

Diabetic foot ulcer (DFU) is a major complication of diabetes and can lead to lower limb amputation if not treated early and properly. In addition to the traditional clinical approaches, in recent years, research on automation using computer vision and machine learning methods plays an important role in DFU classification, achieving promising successes. The most recent automatic approaches to DFU classification are based on convolutional neural networks (CNNs), using solely RGB images as input. In this paper, we present a CNN-based DFU classification method in which we showed that feeding an appropriate feature (texture information) to the CNN model provides a complementary performance to the standard RGB-based deep models of the DFU classification task, and better performance can be obtained if both RGB images and their texture features are combined and used as input to the CNN. To this end, the proposed method consists of two main stages. The first stage extracts texture information from the RGB image using the mapped binary patterns technique. The obtained mapped image is used to aid the second stage in recognizing DFU as it contains texture information of ulcer. The stack of RGB and mapped binary patterns images are fed to the CNN as a tensor input or as a fused image, which is a linear combination of RGB and mapped binary patterns images. The performance of the proposed approach was evaluated using two recently published DFU datasets: the Part-A dataset of healthy and unhealthy (DFU) cases [17] and Part-B dataset of ischaemia and infection cases [18]. The results showed that the proposed methods provided better performance than the state-of-the-art CNN-based methods with 0.981% (AUC) and 0.952% (F-Measure) on the Part-A dataset, 0.995% (AUC) and 0.990% (F-measure) for the Part-B ischaemia dataset, and 0.820% (AUC) and 0.744% (F-measure) on the Part-B infection dataset.

Objective It remains unclear how the different features of radiographic hip osteoarthritis (rHOA) contribute to hip pain. We examined the relationship between rHOA, including its individual components, and hip pain using a novel dual-energy x-ray absorptiometry (DXA)-based method. Methods Hip DXAs were obtained from UK Biobank. A novel automated method obtained minimum joint space width (mJSW) from points placed around the femoral head and acetabulum. Osteophyte areas at the lateral acetabulum, superior and inferior femoral head were derived manually. Semi-quantitative measures of osteophytes and joint space narrowing (JSN) were combined to define rHOA. Logistic regression was used to examine the relationships between these variables and hip pain, obtained via questionnaires. Results 6807 hip DXAs were examined. rHOA was present in 353 (5.2%) individuals and was associated with hip pain [OR 2.42 (1.78–3.29)] and hospital diagnosed OA [6.01 (2.98–12.16)]. Total osteophyte area but not mJSW was associated with hip pain in mutually adjusted models [1.31 (1.23–1.39), 0.95 (0.87–1.04) respectively]. On the other hand, JSN as a categorical variable showed weak associations between grade ≥ 1 and grade ≥ 2 JSN with hip pain [1.30 (1.06–1.60), 1.80 (1.34–2.42) respectively]. Acetabular, superior and inferior femoral osteophyte areas were all independently associated with hip pain [1.13 (1.06–1.20), 1.13 (1.05–1.24), 1.10 (1.03–1.17) respectively]. Conclusion In this cohort, the relationship between rHOA and prevalent hip pain was explained by 2-dimensional osteophyte area, but not by the apparent mJSW. Osteophytes at different locations showed important, potentially independent, associations with hip pain, possibly reflecting the contribution of distinct biomechanical pathways., Highlights • Radiographic hip OA, derived from DXA images, is associated with hip pain in UK Biobank. • Osteophyte size, as reflected by osteophyte area on the DXA image, shows strong associations with hip pain. • In contrast, joint space width is only weakly associated with hip pain.

We present a system that uses convolutional neural networks (CNNs) to detect wrist fractures (distal radius fractures) in posterioanterior and lateral radiographs. The proposed system uses random forest regression voting constrained local model to automatically segment the radius. The resulting automatic annotation is used to register the object across the dataset and crop patches. A CNN is trained on the registered patches for each view separately. Our automatic system outperformed existing systems with a performance of 96% (area under receiver operating characteristic curve) on cross-validation experiments on a dataset of 1010 patients, half of them with fractures.

We describe a fully-automated system for analysing X-rays of the wrist to identify possible fractures. Fractures of the distal radius in the wrist are estimated to be about 18% of the fractures seen in adults and 25% of those seen in children. Unfortunately such fractures are amongst the most frequently missed by doctors in Emergency Departments (EDs). A system which can identify suspicious areas could reduce the number of misdiagnoses. We automatically locate the outline of the radius in both posteroanterior (PA) and lateral (LAT) radiographs, then use shape and texture features to classify abnormalities. We show for the first time that fractures can be better identified in the lateral view, and that combining information from both views leads to an overall improvement in performance.

The commonest diagnostic error in Accident and Emergency (A&E) units is that of missing fractures visible in X-ray images, usually because the doctors are inexperienced or not sufficiently expert. The most commonly missed are wrist fractures [7, 11]. We are developing a fully-automated system for analysing X-rays of the wrist to identify fractures, with the goal of providing prompts to doctors to minimise the number of fractures that are missed. The system automatically locates the outline of the bones (the radius and ulna), then uses shape and texture features to classify abnormalities. The system has been trained and tested on a set of 409 clinical posteroanterior (PA) radiographs of the wrist gathered from a local A&E unit, 199 of which contain fractures. When using the manual shape annotations the system achieves classification performance of 95.5 % (area under the Receiver Operating Characteristic (ROC) curve in cross validation experiments). In fully automatic mode the performance is 88.6 %. Overall the system demonstrates the potential to reduce diagnostic mistakes in A&E.