Detecting hip osteoarthritis on clinical CT: a deep learning application based on 2-D summation images derived from CT.

We developed and compared deep learning models to detect hip osteoarthritis on clinical CT. The CT-based summation images, CT-AP, that resemble X-ray radiographs can detect radiographic hip osteoarthritis and in the absence of large training data, a reliable deep learning model can be optimized by combining CT-AP and X-ray images. INTRODUCTION: In this study, we aimed to investigate the applicability of deep learning (DL) to assess radiographic hip osteoarthritis (rHOA) on computed tomography (CT). METHODS: The study data consisted of 94 abdominopelvic clinical CTs and 5659 hip X-ray images collected from Cohort Hip and Cohort Knee (CHECK). The CT slices were sequentially summed to create radiograph-like 2-D images named CT-AP. X-ray and CT-AP images were classified as rHOA if they had osteoarthritic changes corresponding to Kellgren-Lawrence grade 2 or higher. The study data was split into 55% training, 30% validation, and 15% test sets. A pretrained ResNet18 was optimized for a classification task of rHOA vs. no-rHOA. Five models were trained using (1) X-rays, (2) downsampled X-rays, (3) combination of CT-AP and X-ray images, (4) combination of CT-AP and downsampled X-ray images, and (5) CT-AP images. RESULTS: Amongst the five models, Model-3 and Model-5 performed best in detecting rHOA from the CT-AP images. Model-3 detected rHOA on the test set of CT-AP images with a balanced accuracy of 82.2% and was able to discriminate rHOA from no-rHOA with an area under the receiver operating characteristic curve (ROC AUC) of 0.93 [0.75-0.99]. Model-5 detected rHOA on the test set at a balanced accuracy of 82.2% and classified rHOA from no-rHOA with an ROC AUC of 0.89 [0.67-0.97]. CONCLUSION: CT-based summation images that resemble radiographs can be used to detect rHOA. In addition, in the absence of large training data, a reliable DL model can be optimized by combining CT-AP and X-ray images.

Effect of progressive high-impact exercise on femoral neck structural strength in postmenopausal women with mild knee osteoarthritis: a 12-month RCT.

It is uncertain whether subjects with mild knee osteoarthritis, and who may be at risk of osteoporosis, can exercise safely with the aim of improving hip bone strength. This RCT showed that participating in a high-impact exercise program improved femoral neck strength without any detrimental effects on knee cartilage composition. INTRODUCTION: No previous studies have examined whether high-impact exercise can improve bone strength and articular cartilage quality in subjects with mild knee osteoarthritis. In this 12-month RCT, we assessed the effects of progressive high-impact exercise on femoral neck structural strength and biochemical composition of knee cartilage in postmenopausal women. METHODS: Eighty postmenopausal women with mild knee radiographic osteoarthritis were randomly assigned into the exercise (n = 40) or control (n = 40) group. Femoral neck structural strength was assessed with dual-energy X-ray absorptiometry. The knee cartilage region exposed to exercise loading was measured by the quantitative MRI techniques of T2 mapping and delayed gadolinium-enhanced MRI of cartilage (dGEMRIC). Also, an accelerometer-based body movement monitor was used to evaluate the total physical activity loading on the changes of femoral neck strength in all participants. Training effects on the outcome variables were estimated by the bootstrap analysis of covariance. RESULTS: A significant between-group difference in femoral neck bending strength in favor of the trainees was observed after the 12-month intervention (4.4%, p < 0.01). The change in femoral neck bending strength remained significant after adjusting for baseline value, age, height, and body mass (4.0%, p = 0.020). In all participants, the change in bending strength was associated with the total physical activity loading (r = 0.29, p = 0.012). The exercise participation had no effect on knee cartilage composition. CONCLUSION: The high-impact training increased femoral neck strength without having any harmful effect on knee cartilage in women with mild knee osteoarthritis. These findings imply that progressive high-impact exercise is a feasible method in seeking to prevent hip fractures in postmenopausal women whose articular cartilage may also be frail.

Bone and cartilage characteristics in postmenopausal women with mild knee radiographic osteoarthritis and those without radiographic osteoarthritis.

OBJECTIVES: To evaluate the association between radiographically-assessed knee osteoarthritis and femoral neck bone characteristics in women with mild knee radiographic osteoarthritis and those without radiographic osteoarthritis. METHODS: Ninety postmenopausal women (mean age [SD], 58 [4] years; height, 163 [6] cm; weight, 71 [11] kg) participated in this cross-sectional study. The severity of radiographic knee osteoarthritis was defined using Kellgren-Lawrence grades 0=normal (n=12), 1=doubtful (n=25) or 2=minimal (n=53). Femoral neck bone mineral content (BMC), section modulus (Z), and cross-sectional area (CSA) were measured with DXA. The biochemical composition of ipsilateral knee cartilage was estimated using quantitative MRI measures, T2 mapping and dGEMRIC. The associations between radiographic knee osteoarthritis grades and bone and cartilage characteristics were analyzed using generalized linear models. RESULTS: Age-, height-, and weight-adjusted femoral neck BMC (p for linearity=0.019), Z (p for linearity=0.033), and CSA (p for linearity=0.019) increased significantly with higher knee osteoarthritis grades. There was no linear relationship between osteoarthritis grades and knee cartilage indices. CONCLUSIONS: Increased DXA assessed hip bone strength is related to knee osteoarthritis severity. These results are hypothesis driven that there is an inverse relationship between osteoarthritis and osteoporosis. However, MRI assessed measures of cartilage do not discriminate mild radiographic osteoarthritis severity.

Light physical activity determined by a motion sensor decreases insulin resistance, improves lipid homeostasis and reduces visceral fat in high-risk subjects: PreDiabEx study RCT.

OBJECTIVE: To examine physical activity (PA) thresholds affecting glucose, insulin and lipid concentrations and body fat composition in high-risk patients for type 2 diabetes (T2D). INTERVENTION: A total of 113 subjects of both genders having abnormal glucose levels in the oral glucose tolerance test were contacted. A total of 78 subjects with age 58.8±10.4 years and body mass index 31.7±5.3 kg m(-2) were randomly assigned to intervention and control groups. INTERVENTION consisted of a supervised walking (60 min three times weekly) for 3 months. All the subjects received standard care for PA and weight reduction and wore an accelerometer during the whole wakeful time. RESULTS: Over 80% of the daily steps clustered at an acceleration level of 0.3-0.7 g (2-3 km h(-1) of walking) and were 5870 in the intervention and 4434 in the control group (P<0.029). Between 0 and 3 months no significant changes were observed in fasting and 2-h glucose, body weight or maximal oxygen uptake. In contrast, changes in fasting and 2-h insulin (-3.4 mU l(-1), P=0.035 and -26.6, P=0.003, respectively), homeostasis model assessment-estimated insulin resistance (-1.0, P=0.036), total cholesterol (-0.55 mmol l(-1), P=0.041), low-density lipoprotein (LDL) cholesterol (-0.36 mmol l(-1), P=0.008) and visceral fat area (-5.5 cm(2), P=0.030) were significantly greater in the intervention than in control subjects. The overall effects of PA were analyzed by quartiles of daily steps of all subjects. There were significant reductions in total and LDL cholesterol and visceral fat area between the highest (daily steps over 6520) and the lowest quartile (1780-2810 daily steps). The changes associated with PA remained significant after adjustments of baseline, sex, age and body weight change. CONCLUSION: Habitual and structured PAs with the acceleration levels of 0.3-0.7 g and daily steps over 6520, equivalent to walking at 2-3 km h(-1) for 90 min daily, standing for the relative PA intensity of 30-35% of the maximal oxygen uptake, are clinically beneficial for overweight/obese and physically inactive individuals with a high risk for T2D.

Quantification of differences in bone texture from plain radiographs in knees with and without osteoarthritis.

OBJECTIVE: To quantify differences in bone texture between subjects with different stages of knee osteoarthritis (OA) and age- and gender-matched controls from plain radiographs using advanced image analysis methods. DESIGN: Altogether 203 knees were imaged using constant X-ray parameters and graded according to Kellgren-Lawrence (KL) grading scale (KL0: n = 110, KL1: n = 28, KL2: n = 27, KL3: n = 31, KL4: n = 7). Bone density-related and structure-related parameters were calculated from medial and lateral tibial subchondral bone plate and trabecular bone and from femur. Density-related parameters were derived from grayscale values and structure-related parameters from Laplacian- and local binary patterns (LBP)-based images. RESULTS: Reproducibilities of structure-related parameters were better than bone density-related parameters. Bone density-related parameters were significantly (P < 0.05) higher in KL2-4 groups than in control group (KL0) in medial tibial subchondral bone plate and trabecular bone. LBP-based structure parameters differed significantly between KL0 and KL2-4 groups in medial subchondral bone plate, between KL0 and KL1-4 groups in medial and lateral trabecular bone, and between KL0 and KL1-4/KL2-4 in medial and lateral femur. Laplacian-based parameters differed significantly between KL0 and KL2-4 groups in medial side regions-of-interest (ROIs). CONCLUSIONS: Our results indicate that the changes in bone texture in knee OA can be quantitatively evaluated from plain radiographs using advanced image analysis. Based on the results, increased bone density can be directly estimated if the X-ray imaging conditions are constant between patients. However, structural analysis of bone was more reproducible than direct evaluation of grayscale values, and is therefore better suited for quantitative analysis when imaging conditions are variable.

Discrimination of fractures by low-frequency axial transmission ultrasound in postmenopausal females.

SUMMARY: In this cross-sectional study, 95 postmenopausal women, with and without fracture history, were measured by low-frequency axial transmission ultrasound. The measured ultrasound velocity discriminated the fractured subjects from the nonfractured ones equally or better than peripheral quantitative computed tomography (pQCT) and dual energy x-ray absorptiometry (DXA). These results suggest that low-frequency ultrasound is suitable for bone fragility assessment. INTRODUCTION: Quantitative low-frequency axial transmission ultrasound is a promising modality for assessing mineral density and geometrical properties of long bones such as radius and tibia. The aim of the current study was to evaluate the ability of low-frequency axial transmission ultrasound to discriminate fractures retrospectively in postmenopausal women. METHODS: A cross-sectional study involved 95 female subjects aged 45-88 years, whose fracture information was gathered retrospectively. The fracture group was defined as subjects with one or more low-/moderate-energy fractures. The radius and tibial shaft were measured with a custom-made ultrasonometer to assess the velocity of the low-frequency first-arriving signal (V (LF)). Site-matched pQCT was used to measure volumetric cortical and subcortical bone mineral density (sBMD), and cortical thickness (CTh). Areal BMD (aBMD) was measured using DXA for the whole body (WB), lumbar spine, and hip. RESULTS: The majority (19/32; 59 %) of the fractures were in the upper limb. V (LF) in the radius, but not in the tibia, discriminated fractures with an age- and BMI-adjusted odds ratio (OR) of 2.06 (95 % CI 1.21-3.50, p < 0.01). In the radius, CTh and cortical BMD (CBMD) significantly discriminated fractures, as did the total, cortical, and sBMD in the tibia (adjusted OR 1.35-2.15, p < 0.05). Sensitivity and specificity were similar among all the measurements (area under the receiver operating characteristic curve 0.74-0.81, p < 0.001). CONCLUSIONS: Low-frequency axial transmission ultrasound in the radius was able to discriminate fractured subjects from the nonfractured ones. This suggests that low-frequency axial transmission ultrasound has the potential to assess bone fragility in postmenopausal women.

BMD T-score discriminates trochanteric fractures from unfractured controls, whereas geometry discriminates cervical fracture cases from unfractured controls of similar BMD.

UNLABELLED: The ability of bone mineral density (BMD) to discriminate cervical and trochanteric hip fractures was studied. Since the majority of fractures occur among people who are not diagnosed as having osteoporosis, we also examined this population to elucidate whether geometrical risk factors can yield additional information on hip fracture risk beside BMD. The study showed that the T-score criterion was able to discriminate fracture patients from controls in the cases of trochanteric fractures, whereas geometrical measures may discriminate cervical fracture cases in patients with T-score >-2.5. INTRODUCTION: Low bone mineral density (BMD) is a well-established risk factor for hip fracture. However, majority of fractures occur among people not diagnosed as having osteoporosis. We studied the ability of BMD to discriminate cervical and trochanteric hip fractures. Furthermore, we examined whether geometrical measures can yield additional information on the assessment of hip fracture risk in the fracture cases in subjects with T-score >-2.5. METHODS: Study group consisted of postmenopausal females with non-pathologic cervical (n = 39) or trochanteric (n = 18) hip fracture (mean age 74.2 years) and 40 age-matched controls. BMD was measured at femoral neck, and femoral neck axis length, femoral neck and shaft cortex thicknesses (FNC and FSC), and femoral neck-shaft angle (NSA) were measured from radiographs. RESULTS: BMD T-score threshold of -2.5 was able to discriminate trochanteric fractures from controls (p < 0.001). Seventeen out of 18 trochanteric fractures occurred in individuals with T-score -2.5. Within these non-osteoporotic cervical fracture patients (N = 19) and non-osteoporotic controls (N = 35), 83.3% were classified correctly based on a model including NSA and FNC (p < 0.001), area under the receiver operating characteristics curve being 0.85 for the model, while it was only 0.56 for BMD alone. CONCLUSIONS: The study suggests that the risk of trochanteric fractures could be discriminated based on a BMD T-score <-2.5 criterion, whereas cervical fracture cases would remain under-diagnosed if solely using this criterion. Instead, geometrical risk factors are able to discriminate cervical fracture cases even among individuals with T-score >-2.5. For cervical and trochanteric fractures combined, BMD and geometric measures independently contributed to hip fracture discrimination. Our data support changing from T-score <-2.5 to a more comprehensive assessment of hip fracture etiology, in which fracture type is also taken into account. The findings need to be confirmed with a larger sample, preferably in a prospective study.

Effect of impact exercise on bone metabolism.

SUMMARY: Regular impact exercise in premenopausal women caused positive osteogenic effects associated to low basal serum parathormone (PTH) but had no effects on bone turnover markers PINP or TRACP5b. The low serum basal PTH levels during impact exercise may be a sign of increased incorporation of calcium to bone. INTRODUCTION: This study aimed to determine the long-term effects of high-impact exercise on bone turnover and calciotropic hormones. METHODS: We performed a 12-month population-based, randomized, controlled exercise trial in 120 women (age 35-40 years) randomly assigned to an exercise group (EG; n = 60) or a control group (CG; n = 60). The exercise regimen consisted of supervised high-impact exercises three times per week. Daily impact loading was assessed by using an accelerometer. Bone turnover markers and calciotropic hormones were analyzed at 0, 6, and 12 months. RESULTS: Twelve months of impact exercise did not reveal any treatment effects in bone turnover markers PINP or TRAPC5b, whereas serum basal PTH decreased significantly more in the EG than in the CG (-11.2 vs. -2.2 pg/mL; p = 0.03). The change in PTH was dose dependent and most clearly seen in subjects with 96 to 130 daily impacts at 2.5 to 5.3 g (e.g., running or jumping). CONCLUSIONS: Regular impact exercise does not cause persistent alterations in bone turnover emphasizing necessity of continuous training to achieve bone benefits. Impact exercise training lowers the serum basal PTH levels and possibly enables greater difference between the basal PTH and transient exercise-induced PTH peaks leading to osteogenic effects.

High insulin levels are positively associated with peripheral nervous system function.

OBJECTIVE: The aim of this study was to analyze peripheral nervous system (PNS) function in overweight and obese individuals. MATERIALS AND METHODS: Forty-four adult non-diabetic overweight individuals were recruited. Peroneal motor nerve conduction and radial, sural, and medial plantar sensory nerve conduction were studied. Insulin and glucose levels were determined twice (over a 2- to 3-year period) with an oral glucose tolerance test (OGTT). Multiple stepwise linear regression models adjusted for age, height, weight, and skin temperature were used to analyze the data. RESULTS: Analysis revealed that baseline insulin levels measured 120 min after an OGTT explained 18% of the variation in peroneal F-wave minimum latency, 8% of peroneal F-wave maximum latency variation, 15% of sural sensory latency variation, 13% of sural sensory nerve conduction velocity (NCV) variation, and 10% of the variation in medial plantar sensory NCV. DISCUSSION AND CONCLUSION: Our study shows that serum insulin levels measured 120 min after an OGGT are positively associated with PNS function. High insulin levels without notably high glucose levels appear to be beneficial for the function of the PNS.

Are signs of temporomandibular disorders stable and predictable in adolescents with headache?

The aim of the present study was to study changes in signs and symptoms of temporomandibular disorders (TMD) and factors predicting TMD signs in adolescents with and without headache. A population-based sample (n = 212) of 13-year-olds with and without headache was re-examined at the age of 16. The study included a questionnaire, face-to-face interview and somatic examination. In addition, a neurological examination, a muscle evaluation and a stomatognathic examination were performed. Significant changes were seen in TMD signs during the follow-up, but TMD signs at the end of the follow-up could not be predicted by baseline headache, sleeping difficulties, depression or muscle pain. TMD signs at the age of 16 were associated with female gender and muscle pain. We conclude that considerable changes in TMD signs occur in the follow-up of adolescents with and without headache. Headache-related TMD are not predictable in adolescents with and without headache.