A Preliminary Study of Quantitative MRI Cartilage Loss Fraction and Its Association With Future Arthroplasty Using the Osteoarthritis Initiative Database.

Background and objective  Osteoarthritis (OA) is the most common arthritis in the world. Despite the high disease burden, there is no therapy to prevent, halt, or reverse OA, and many clinical trials relied on radiographic biomarkers for therapy response. It is important to identify patients with early OA who will eventually need arthroplasty, the end-stage treatment for osteoarthritis. This pilot study evaluates a novel MRI biomarker, cartilage loss fraction, for association with future arthroplasty and evaluates its feasibility of use and effect size estimates. Materials and methods Publicly available knee MRIs from the Osteoarthritis Initiative were used. A total of 38 participants with Kellgren-Lawrence (K-L) grade >1 and 38 participants with K-L grade ≤ 1 at enrollment were matched in age, sex, race, and BMI, and assessed for the degree of full-thickness cartilage loss, or cartilage loss fraction. Univariate conditional logistic regression analysis was performed for differences in cartilage loss fractions between groups. Receiver operating characteristic (ROC) curve analysis was performed to assess the association of MRI biomarkers and knee arthroplasty during the eight-year follow-up. Results The medial femoral condyle, medial tibial plateau, total, and two-year progression cartilage loss fractions were significantly higher in participants with K-L grade >1 (p < 0.01 for all) and showed high area under the curve (AUC) values on ROC analysis (812, 0.827, 0.917, and 0.933, respectively). These results were comparable or more strongly associated with other OA grading schemes. Conclusion MRI biomarker cartilage loss fractions are significantly higher in subjects with K-L grade >1 and show a strong association with arthroplasty. After further validation, cartilage loss fracture may be used to predict future arthroplasty.

Opportunistic Screening for Osteoporosis Using Hand Radiographs - A Feature Augmentation Study Technique (FAST).

Patients with low bone mineral density (BMD) face an increased risk of fractures, yet are frequently undiagnosed. Consequently, it is imperative to have opportunistically screen for low BMD in patients undergoing other medical evaluations. This retrospective study encompassed 422 patients aged ≥ 50 who underwent both dual-energy X-ray absorptiometry (DXA) and hand radiographs (modality of digital X-ray) from three different vendors within a 12-month period. The dataset was randomly divided into training/validation (n=338) and test (n=84) datasets. we sought to predict osteoporosis/osteopenia and establish correlations between bone textural analysis and DXA measurements. Our results demonstrate that the deep learning model achieved an accuracy of 77.38%, sensitivity of 77.38%, specificity of 73.63%, and an area under the curve (AUC) of 83% in detecting osteoporosis/osteopenia. These findings suggest that hand radiographs can serve as a viable screening tool for identifying individuals warranting formal DXA assessment for osteoporosis/osteopenia.

Utilizing Generalized Linear Mixed Effects Models for Prediction of Bone Mineral Decrease After Lung Transplant - A Preliminary Study.

INTRODUCTION: Recipients of lung transplants are at heightened risk of developing low bone mineral density (BMD). They also have increased risk of developing osteoporosis-related fractures. This rate of BMD decline is not well characterized. The aim of this manuscript is to characterize the rate of BMD decrease after lung transplant. METHODOLOGY: This is a preliminary retrospective cohort study of 200 patients who received lung transplants. Each patient had pre-transplant dual energy X-ray absorptiometry (DXA) and post-transplant DXA scans on the same DXA scanner. The BMD at each lumbar vertebra, total hip, and femoral neck were recorded. Generalized linear mixed effects models with random intercepts and random slopes were used to model the rate of change of BMD after transplant, adjusting for time since transplant, sex, and sex*time interaction. RESULTS: We found that men had higher baseline BMD than women at all sites (P<0.05 for all). The rate of BMD decrease was fastest at the femoral neck (P<0.05). Men lost BMD at a faster rate (-5.23 x10-5 g/cm2/day) than women at the femoral neck (-2.22 x10-5 g/cm2 /day). CONCLUSION: BMD decreases after lung transplant and occurs faster in men than women.

Can we screen opportunistically for low bone mineral density using CT scans of the shoulder and artificial intelligence?

OBJECTIVE: To evaluate whether the CT attenuation of bones seen on shoulder CT scans could be used to predict low bone mineral density (BMD) (osteopenia/osteoporosis), and to compare the performance of two machine learning models to predict low BMD. METHODS: In this study, we evaluated 194 patients aged 50 years or greater (69.2 ± 9.1 years; 170 females) who underwent unenhanced shoulder CT scans and dual-energy X-ray absorptiometry within 1 year of each other between January 1, 2010, and December 31, 2021. The CT attenuation of the humerus, glenoid, coracoid, acromion, clavicle, first, second, and third ribs was obtained using 3D-Slicer. Support vector machines (SVMs) and k-nearest neighbours (kNN) were used to predict low BMD. DeLong test was used to compare the areas under the curve (AUCs). RESULTS: A CT attenuation of 195.4 Hounsfield Units of the clavicle had a sensitivity of 0.577, specificity of 0.781, and AUC of 0.701 to predict low BMD. In the test dataset, the SVM had sensitivity of 0.686, specificity of 1.00, and AUC of 0.857, while the kNN model had sensitivity of 0.966, specificity of 0.200, and AUC of 0.583. The SVM was superior to the CT attenuation of the clavicle (P = .003) but not better than the kNN model (P = .098). CONCLUSION: The CT attenuation of the clavicle was best for predicting low BMD; however, a multivariable SVM was superior for predicting low BMD. ADVANCES IN KNOWLEDGE: SVM utilizing the CT attenuations at many sites was best for predicting low BMD.

Phenotypic and ancestry-related assortative mating in autism.

BACKGROUND: Positive assortative mating (AM) in several neuropsychiatric traits, including autism, has been noted. However, it is unknown whether the pattern of AM is different in phenotypically defined autism subgroups [e.g., autism with and without intellectually disability (ID)]. It is also unclear what proportion of the phenotypic AM can be explained by the genetic similarity between parents of children with an autism diagnosis, and the consequences of AM on the genetic structure of the population. METHODS: To address these questions, we analyzed two family-based autism collections: the Simons Foundation Powering Autism Research for Knowledge (SPARK) (1575 families) and the Simons Simplex Collection (SSC) (2283 families). RESULTS: We found a similar degree of phenotypic and ancestry-related AM in parents of children with an autism diagnosis regardless of the presence of ID. We did not find evidence of AM for autism based on autism polygenic scores (PGS) (at a threshold of |r|> 0.1). The adjustment of ancestry-related AM or autism PGS accounted for only 0.3-4% of the fractional change in the estimate of the phenotypic AM. The ancestry-related AM introduced higher long-range linkage disequilibrium (LD) between single nucleotide polymorphisms (SNPs) on different chromosomes that are highly ancestry-informative compared to SNPs that are less ancestry-informative (D2 on the order of 1 × 10-5). LIMITATIONS: We only analyzed participants of European ancestry, limiting the generalizability of our results to individuals of non-European ancestry. SPARK and SSC were both multicenter studies. Therefore, there could be ancestry-related AM in SPARK and SSC due to geographic stratification. The study participants from each site were unknown, so we were unable to evaluate for geographic stratification. CONCLUSIONS: This study showed similar patterns of AM in autism with and without ID, and demonstrated that the common genetic influences of autism are likely relevant to both autism groups. The adjustment of ancestry-related AM and autism PGS accounted for < 5% of the fractional change in the estimate of the phenotypic AM. Future studies are needed to evaluate if the small increase of long-range LD induced by ancestry-related AM has impact on the downstream analysis.

Sleep and Activity Patterns in Autism Spectrum Disorder.

Background: Autism spectrum disorder (ASD) is a highly heritable and heterogeneous neurodevelopmental disorder characterized by impaired social interactions, repetitive behaviors, and a wide range of comorbidities. Between 44-83% of autistic individuals report sleep disturbances, which may share an underlying neurodevelopmental basis with ASD. Methods: We recruited 382 ASD individuals and 223 of their family members to obtain quantitative ASD-related traits and wearable device-based accelerometer data spanning three consecutive weeks. An unbiased approach identifying traits associated with ASD was achieved by applying the elastic net machine learning algorithm with five-fold cross-validation on 6,878 days of data. The relationship between sleep and physical activity traits was examined through linear mixed-effects regressions using each night of data. Results: This analysis yielded 59 out of 242 actimetry measures associated with ASD status in the training set, which were validated in a test set (AUC: 0.777). For several of these traits (e.g. total light physical activity), the day-to-day variability, in addition to the mean, was associated with ASD. Individuals with ASD were found to have a stronger correlation between physical activity and sleep, where less physical activity decreased their sleep more significantly than that of their non-ASD relatives. Conclusions: The average duration of sleep/physical activity and the variation in the average duration of sleep/physical activity strongly predict ASD status. Physical activity measures were correlated with sleep quality, traits, and regularity, with ASD individuals having stronger correlations. Interventional studies are warranted to investigate whether improvements in both sleep and increased physical activity may improve the core symptoms of ASD.

Advancing Diagnostics and Patient Care: The Role of Biomarkers in Radiology.

The integration of biomarkers into medical practice has revolutionized the field of radiology, allowing for enhanced diagnostic accuracy, personalized treatment strategies, and improved patient care outcomes. This review offers radiologists a comprehensive understanding of the diverse applications of biomarkers in medicine. By elucidating the fundamental concepts, challenges, and recent advancements in biomarker utilization, it will serve as a bridge between the disciplines of radiology and epidemiology. Through an exploration of various biomarker types, such as imaging biomarkers, molecular biomarkers, and genetic markers, I outline their roles in disease detection, prognosis prediction, and therapeutic monitoring. I also discuss the significance of robust study designs, blinding, power and sample size calculations, performance metrics, and statistical methodologies in biomarker research. By fostering collaboration between radiologists, statisticians, and epidemiologists, I hope to accelerate the translation of biomarker discoveries into clinical practice, ultimately leading to improved patient care.

Intravenous tranexamic acid is associated with an increased risk of pulmonary embolism following sarcoma resection.

INTRODUCTION: Tranexamic acid (TXA) is an antifibrinolytic drug that has been shown to reduce blood loss following surgery. The use of TXA during orthopedic procedures has gained widespread acceptance, with multiple clinical studies demonstrating no increase in thrombotic complications. While TXA has been shown to be safe and effective for several orthopedic procedures, its use in orthopedic sarcoma surgery is not well established. Cancer-associated thrombosis remains a significant cause of morbidity and mortality in patients with sarcoma. It is unknown if intraoperative TXA use will increase the risk of developing a postoperative thrombotic complication in this population. This study aimed to compare the risk of postoperative thrombotic complications in patients who received TXA during sarcoma resection to patients who did not receive TXA. METHODS: A retrospective review was performed of 1099 patients who underwent resection of a soft tissue or bone sarcoma at our institution between 2010 and 2021. Baseline demographics and postoperative outcomes were compared between patients who did and did not receive intraoperative TXA. We evaluated 90-day complication rates, including: deep venous thrombosis (DVT), pulmonary embolism (PE), myocardial infarction (MI), cerebrovascular accident (CVA), and mortality. RESULTS: TXA was used more commonly for bone tumors (p < 0.001), tumors located in the pelvis (p = 0.004), and larger tumors (p < 0.001). Patients who received intraoperative TXA were associated with a significant increase in developing a postoperative DVT (odds ratio [OR]: 2.22, p = 0.036) and PE (OR: 4.62, p < 0.001), but had no increase in CVA, MI, or mortality (all p > 0.05) within 90 days of surgery, following univariate analysis. Multivariable analysis confirmed that TXA was independently associated with developing a postoperative PE (OR: 10.64, 95% confidence interval: 2.23-50.86, p = 0.003). We found no association with DVT, MI, CVA, or mortality within 90 days postoperatively, following intraoperative TXA use. CONCLUSION: Our results demonstrate a higher associated risk of PE following TXA use in sarcoma surgery and caution is warranted with TXA use in this patient population.

Automated detection of causal relationships among diseases and imaging findings in textual radiology reports.

OBJECTIVE: Textual radiology reports contain a wealth of information that may help understand associations among diseases and imaging observations. This study evaluated the ability to detect causal associations among diseases and imaging findings from their co-occurrence in radiology reports. MATERIALS AND METHODS: This IRB-approved and HIPAA-compliant study analyzed 1 702 462 consecutive reports of 1 396 293 patients; patient consent was waived. Reports were analyzed for positive mention of 16 839 entities (disorders and imaging findings) of the Radiology Gamuts Ontology (RGO). Entities that occurred in fewer than 25 patients were excluded. A Bayesian network structure-learning algorithm was applied at P < 0.05 threshold: edges were evaluated as possible causal relationships. RGO and/or physician consensus served as ground truth. RESULTS: 2742 of 16 839 RGO entities were included, 53 849 patients (3.9%) had at least one included entity. The algorithm identified 725 pairs of entities as causally related; 634 were confirmed by reference to RGO or physician review (87% precision). As shown by its positive likelihood ratio, the algorithm increased detection of causally associated entities 6876-fold. DISCUSSION: Causal relationships among diseases and imaging findings can be detected with high precision from textual radiology reports. CONCLUSION: This approach finds causal relationships among diseases and imaging findings with high precision from textual radiology reports, despite the fact that causally related entities represent only 0.039% of all pairs of entities. Applying this approach to larger report text corpora may help detect unspecified or heretofore unrecognized associations.

A snapshot of artificial intelligence research 2019-2021: is it replacing or assisting physicians?

Artificial intelligence (AI) has the potential to be a disruptive technology in healthcare. Recently, there is increased speculation that AI may be used to replace healthcare providers in the future. To answer this question, we reviewed over 21 000 articles published in medical specialty journals between 2019 and 2021 to evaluate whether these AI models were intended to assist or replace healthcare providers. We also evaluated whether all Food and Drug Administration (FDA)-approved AI models were used to assist or replace healthcare providers. We find that most AI models published in this time period were intended to assist rather than replace healthcare providers, and that most of the published AI models performed tasks that could not be done by healthcare providers.

Detecting whether L1 or other lumbar levels would be excluded from DXA bone mineral density analysis during opportunistic CT screening for osteoporosis using machine learning.

PURPOSE: One or more vertebrae are sometimes excluded from dual-energy X-ray absorptiometry (DXA) analysis if the bone mineral density (BMD) T-score estimates are not consistent with the other lumbar vertebrae BMD T-score estimates. The goal of this study was to build a machine learning framework to identify which vertebrae would be excluded from DXA analysis based on the computed tomography (CT) attenuation of the vertebrae. METHODS: Retrospective review of 995 patients (69.0% female) aged 50 years or greater with CT scans of the abdomen/pelvis and DXA within 1 year of each other. Volumetric semi-automated segmentation of each vertebral body was performed using 3D-Slicer to obtain the CT attenuation of each vertebra. Radiomic features based on the CT attenuation of the lumbar vertebrae were created. The data were randomly split into training/validation (90%) and test datasets (10%). We used two multivariate machine learning models: a support vector machine (SVM) and a neural net (NN) to predict which vertebra(e) were excluded from DXA analysis. RESULTS: L1, L2, L3, and L4 were excluded from DXA in 8.7% (87/995), 9.9% (99/995), 32.3% (321/995), and 42.6% (424/995) patients, respectively. The SVM had a higher area under the curve (AUC = 0.803) than the NN (AUC = 0.589) for predicting whether L1 would be excluded from DXA analysis (P = 0.015) in the test dataset. The SVM was better than the NN for predicting whether L2 (AUC = 0.757 compared to AUC = 0.478), L3 (AUC = 0.699 compared to AUC = 0.555), or L4 (AUC = 0.751 compared to AUC = 0.639) were excluded from DXA analysis. CONCLUSIONS: Machine learning algorithms could be used to identify which lumbar vertebrae would be excluded from DXA analysis and should not be used for opportunistic CT screening analysis. The SVM was better than the NN for identifying which lumbar vertebra should not be used for opportunistic CT screening analysis.

Opportunistic Screening for Osteoporosis Using CT Scans of the Knee: A Pilot Study.

Knee CT scans are used for planning for total knee arthroplasties in patients who are often simultaneously at risk for frailty fractures due to low bone mineral density. We retrospectively identified 200 patients (85.5% female) with concurrent CT scans of the knee and Dual energy x-ray absorptiometry (DXA). The mean CT attenuation of the distal femur, proximal tibia and fibula, and patella, were calculated using volumetric 3-dimensional segmentation using 3D Slicer. Data were split randomly into training 80% and test 20% datasets. The optimal CT attenuation threshold for the proximal fibula was obtained in the training dataset and evaluated in the test dataset. A support vector machine (SVM) with radial basis function (RBF) using C-classification was trained and tuned using 5-fold cross-validation in the training dataset and then evaluated in the test dataset. The SVM had a higher area-under-the curve (AUC) of 0.937 and better performance to detect osteoporosis/osteopenia than the CT attenuation of the fibula (AUC of 0.717) (P=0.015). Opportunistic screening for osteoporosis/osteopenia could be accomplished using CT scans of the knee.

Opportunistic Screening for Osteoporosis Using Hand Radiographs: A Preliminary Study.

Patients with low bone mineral density (BMD) are at risk for fractures however are often undiagnosed. Therefore, there is a need to opportunistically screen for low BMD in patients who present for other studies. This is a retrospective study of 812 patients aged 50 years or older who had dual-energy X-ray absorptiometry (DXA) and radiographs of the hands within 12 months of each other. This dataset was randomly split into training/validation (n=533) and test (n=136) datasets. A deep learning (DL) framework was used to predict osteoporosis/osteopenia. Correlations between the textural analysis of the bones and DXA measurements were obtained. We found that the DL model had an accuracy of 82.00%, sensitivity of 87.03%, specificity of 61.00% and an area under the curve (AUC) of 74.00% to detect osteoporosis/osteopenia. Our findings show that radiographs of the hand can be used to screen for osteoporosis/osteopenia and identify patients who should get formal DXA evaluation.

Direct anterior approach associated with lower dislocation risk after primary total hip arthroplasty in patients with prior lumbar spine fusion.

BACKGROUND: While there has been much interest in the increased dislocation rate in total hip arthroplasty (THA) patients with a lumbar spine fusion (LSF), there is minimal literature comparing the risk based on surgical approach. The purpose of this study was to determine if a direct anterior (DA) approach was protective against dislocation when compared to the anterolateral and posterior approaches in this high-risk patient population. METHODS: A retrospective review was performed of 6554 THAs performed at our institution from January 2011 to May 2021. 294 (4.5%) patients had a prior LSF and were included in the analysis. The surgical approach, timing of LSF in relation to THA, vertebral levels fused, timing of THA dislocation, and the need for revision surgery were recorded for statistical analysis. RESULTS: In total, 39.7.3% of patients underwent a DA approach (n = 117), 25.9% underwent an anterolateral approach (n = 76), and 34.3% underwent a posterior approach (n = 101). There was no difference in number of vertebral levels fused between groups (mean 2.5, all p > 0.05). There was a total of 13 (4.4%) THA dislocation events, with an average time from surgery to dislocation of 5.6 months (0.3-30.5 months). There were fewer dislocations in the DA cohort (0.9%) in comparison to both the anterolateral (6.6%, p = 0.036) and posterior groups (6.9%, p = 0.026). CONCLUSIONS: The DA approach demonstrated a significantly lower THA dislocation rate compared to both the anterolateral and posterior approaches in patients with a concomitant LSF.

Machine Learning for Opportunistic Screening for Osteoporosis and Osteopenia Using Knee CT Scans.

PURPOSE: To predict whether a patient has osteoporosis/osteopenia using the attenuation of trabecular bone obtained from knee computed tomography (CT) scans. METHODS: Retrospective analysis of 273 patients who underwent contemporaneous knee CT scans and dual-energy X-ray absorptiometry (DXA) within 1 year. Volumetric segmentation of the trabecular bone of the distal femur, proximal tibia, patella, and proximal fibula was performed to obtain the bone CT attenuation. The data was randomly split into training/validation (78%) and test (22%) datasets and the performance in the test dataset were evaluated. The predictive properties of the CT attenuation of each bone to predict osteoporosis/osteopenia were assessed. Multivariable support vector machines (SVM) and random forest classifiers (RF) were used to predict osteoporosis/osteopenia. RESULTS: Patients with a mean age (range) of 67.9 (50-87) years, 85% female were evaluated. Seventy-seven (28.2%) of patients had normal bone mineral density (BMD), 140 (51.3%) had osteopenia, and 56 (20.5%) had osteoporosis. The proximal tibia had the best predictive ability of all bones and a CT attenuation threshold of 96.0 Hounsfield Units (HU) had a sensitivity of .791, specificity of .706, and area under the curve (AUC) of .748. The AUC for the SVM with cubic kernel classifier (AUC = .912) was better than the RF classifier (AUC = .683, P < .001) and better than using the CT attenuation threshold of 96.0 HU at the proximal tibia (AUC = .748, P = .025). CONCLUSIONS: Opportunistic screening for osteoporosis/osteopenia can be performed using knee CT scans. Multivariable machine learning models are more predictive than the CT attenuation of a single bone.

Radiomics analysis in medical imaging research.

This article discusses the current research in the field of radiomics in medical imaging with emphasis on its role in fighting coronavirus disease 2019 (COVID-19). This article covers the building of radiomic models in a simple straightforward manner, while discussing radiomic models potential to help us face this pandemic.