Search

Abstract Large language models (LLMs) have broad medical knowledge and can reason about medical information across many domains, holding promising potential for diverse medical applications in the near future. In this study, we demonstrate a concerning vulnerability of LLMs in medicine. Through targeted manipulation of just 1.1% of the weights of the LLM, we can deliberately inject incorrect biomedical facts. The erroneous information is then propagated in the model’s output while maintaining performance on other biomedical tasks. We validate our findings in a set of 1025 incorrect biomedical facts. This peculiar susceptibility raises serious security and trustworthiness concerns for the application of LLMs in healthcare settings. It accentuates the need for robust protective measures, thorough verification mechanisms, and stringent management of access to these models, ensuring their reliable and safe use in medical practice.

Abstract Background Wound healing complications are a major challenge following the extended lateral approach in calcaneal fractures. Soft-tissue microcirculation plays an important role via the delivery of oxygen, nutrients, and the regulation of a local milieu. The aim of this clinical study was to examine the effect of intraoperative pneumatic tourniquet application on skin and subcutaneous microcirculation, and its impact on wound healing progression. Methods Patients with calcaneal fractures were randomly assigned to two groups defined by a surgery conducted either with use or without use of a tourniquet. Blood flow (BF [AU]), tissue oxygen saturation (SO 2 [%]) and the relative amount of haemoglobin (rHb[AU]) were intraoperatively measured at two depths (2 and 8 mm) non-invasively by spectrophotometry (Micro-Lightguide O2C ® , LEA Medizintechnik, Giessen, Germany). Time points were before and after inflation of the pneumatic tourniquet and also at the end of surgery before deflation. A linear mixed model (LMM) was fitted for statistical analysis. Results Thirty-four patients (3 women and 31 men) with 37 calcaneal fractures were included. In 22 of them, the surgery was conducted with a tourniquet and in the other 15 without its use. A significant decrease of microcirculation, characterized by decreases in blood flow (p = 0.011) and tissue oxygenation (p = 0.023) was measured in 8 mm depth after inflating the tourniquet. However, these changes did not influence the time of postoperative wound healing. Conclusion The use of a pneumatic tourniquet reduces deep microcirculation without affecting postoperative wound healing. Trial registration The study was registered in www.ClinicalTrials.gov (NCT01264146).

Abstract Background Quantitative techniques such as T2 and T1ρ mapping allow evaluating the cartilage and meniscus. We evaluated multi-interleaved X-prepared turbo-spin echo with intuitive relaxometry (MIXTURE) sequences with turbo spin-echo (TSE) contrast and additional parameter maps versus reference TSE sequences in an in situ model of human cartilage defects. Methods Standardized cartilage defects of 8, 5, and 3 mm in diameter were created in the lateral femora of ten human cadaveric knee specimens (81 ± 10 years old; nine males, one female). MIXTURE sequences providing proton density-weighted fat-saturated images and T2 maps or T1-weighted images and T1ρ maps as well as the corresponding two- and three-dimensional TSE reference sequences were acquired before and after defect creation (3-T scanner; knee coil). Defect delineability, bone texture, and cartilage relaxation times were quantified. Appropriate parametric or non-parametric tests were used. Results Overall, defect delineability and texture features were not significantly different between the MIXTURE and reference sequences (p ≤ 0.47). After defect creation, relaxation times significantly increased in the central femur (T2pre = 51 ± 4 ms [mean ± standard deviation] versus T2post = 56 ± 4 ms; p = 0.002) and all regions combined (T1ρpre = 40 ± 4 ms versus T1ρpost = 43 ± 4 ms; p = 0.004). Conclusions MIXTURE permitted time-efficient simultaneous morphologic and quantitative joint assessment based on clinical image contrasts. While providing T2 or T1ρ maps in clinically feasible scan time, morphologic image features, i.e., cartilage defects and bone texture, were comparable between MIXTURE and reference sequences. Relevance statement Equally time-efficient and versatile, the MIXTURE sequence platform combines morphologic imaging using familiar contrasts, excellent image correspondence versus corresponding reference sequences and quantitative mapping information, thereby increasing the diagnostic value beyond mere morphology. Key points • Combined morphologic and quantitative MIXTURE sequences are based on three-dimensional TSE contrasts. • MIXTURE sequences were studied in an in situ human cartilage defect model. • Morphologic image features, i.e., defect delineabilty and bone texture, were investigated. • Morphologic image features were similar between MIXTURE and reference sequences. • MIXTURE allowed time-efficient simultaneous morphologic and quantitative knee joint assessment. Graphical Abstract

Summary: Reliably detecting potentially misleading patterns in automated diagnostic assistance systems, such as those powered by artificial intelligence (AI), is crucial for instilling user trust and ensuring reliability. Current techniques fall short in visualizing such confounding factors. We propose DiffChest, a self-conditioned diffusion model trained on 515,704 chest radiographs from 194,956 patients across the US and Europe. DiffChest provides patient-specific explanations and visualizes confounding factors that might mislead the model. The high inter-reader agreement, with Fleiss’ kappa values of 0.8 or higher, validates its capability to identify treatment-related confounders. Confounders are accurately detected with 10%–100% prevalence rates. The pretraining process optimizes the model for relevant imaging information, resulting in excellent diagnostic accuracy for 11 chest conditions, including pleural effusion and heart insufficiency. Our findings highlight the potential of diffusion models in medical image classification, providing insights into confounding factors and enhancing model robustness and reliability.

Abstract Background To compare denoising diffusion probabilistic models (DDPM) and generative adversarial networks (GAN) for recovering contrast-enhanced breast magnetic resonance imaging (MRI) subtraction images from virtual low-dose subtraction images. Methods Retrospective, ethically approved study. DDPM- and GAN-reconstructed single-slice subtraction images of 50 breasts with enhancing lesions were compared to original ones at three dose levels (25%, 10%, 5%) using quantitative measures and radiologic evaluations. Two radiologists stated their preference based on the reconstruction quality and scored the lesion conspicuity as compared to the original, blinded to the model. Fifty lesion-free maximum intensity projections were evaluated for the presence of false-positives. Results were compared between models and dose levels, using generalized linear mixed models. Results At 5% dose, both radiologists preferred the GAN-generated images, whereas at 25% dose, both radiologists preferred the DDPM-generated images. Median lesion conspicuity scores did not differ between GAN and DDPM at 25% dose (5 versus 5, p = 1.000) and 10% dose (4 versus 4, p = 1.000). At 5% dose, both readers assigned higher conspicuity to the GAN than to the DDPM (3 versus 2, p = 0.007). In the lesion-free examinations, DDPM and GAN showed no differences in the false-positive rate at 5% (15% versus 22%), 10% (10% versus 6%), and 25% (6% versus 4%) (p = 1.000). Conclusions Both GAN and DDPM yielded promising results in low-dose image reconstruction. However, neither of them showed superior results over the other model for all dose levels and evaluation metrics. Further development is needed to counteract false-positives. Relevance statement For MRI-based breast cancer screening, reducing the contrast agent dose is desirable. Diffusion probabilistic models and generative adversarial networks were capable of retrospectively enhancing the signal of low-dose images. Hence, they may supplement imaging with reduced doses in the future. Key points • Deep learning may help recover signal in low-dose contrast-enhanced breast MRI. • Two models (DDPM and GAN) were trained at different dose levels. • Radiologists preferred DDPM at 25%, and GAN images at 5% dose. • Lesion conspicuity between DDPM and GAN was similar, except at 5% dose. • GAN and DDPM yield promising results in low-dose image reconstruction. Graphical Abstract

Abstract Background Artificial intelligence (AI) models are increasingly used in the medical domain. However, as medical data is highly sensitive, special precautions to ensure its protection are required. The gold standard for privacy preservation is the introduction of differential privacy (DP) to model training. Prior work indicates that DP has negative implications on model accuracy and fairness, which are unacceptable in medicine and represent a main barrier to the widespread use of privacy-preserving techniques. In this work, we evaluated the effect of privacy-preserving training of AI models regarding accuracy and fairness compared to non-private training. Methods We used two datasets: (1) A large dataset (N = 193,311) of high quality clinical chest radiographs, and (2) a dataset (N = 1625) of 3D abdominal computed tomography (CT) images, with the task of classifying the presence of pancreatic ductal adenocarcinoma (PDAC). Both were retrospectively collected and manually labeled by experienced radiologists. We then compared non-private deep convolutional neural networks (CNNs) and privacy-preserving (DP) models with respect to privacy-utility trade-offs measured as area under the receiver operating characteristic curve (AUROC), and privacy-fairness trade-offs, measured as Pearson’s r or Statistical Parity Difference. Results We find that, while the privacy-preserving training yields lower accuracy, it largely does not amplify discrimination against age, sex or co-morbidity. However, we find an indication that difficult diagnoses and subgroups suffer stronger performance hits in private training. Conclusions Our study shows that – under the challenging realistic circumstances of a real-life clinical dataset – the privacy-preserving training of diagnostic deep learning models is possible with excellent diagnostic accuracy and fairness.

Abstract Background Pretraining labeled datasets, like ImageNet, have become a technical standard in advanced medical image analysis. However, the emergence of self-supervised learning (SSL), which leverages unlabeled data to learn robust features, presents an opportunity to bypass the intensive labeling process. In this study, we explored if SSL for pretraining on non-medical images can be applied to chest radiographs and how it compares to supervised pretraining on non-medical images and on medical images. Methods We utilized a vision transformer and initialized its weights based on the following: (i) SSL pretraining on non-medical images (DINOv2), (ii) supervised learning (SL) pretraining on non-medical images (ImageNet dataset), and (iii) SL pretraining on chest radiographs from the MIMIC-CXR database, the largest labeled public dataset of chest radiographs to date. We tested our approach on over 800,000 chest radiographs from 6 large global datasets, diagnosing more than 20 different imaging findings. Performance was quantified using the area under the receiver operating characteristic curve and evaluated for statistical significance using bootstrapping. Results SSL pretraining on non-medical images not only outperformed ImageNet-based pretraining (p

Abstract A knowledge gap persists between machine learning (ML) developers (e.g., data scientists) and practitioners (e.g., clinicians), hampering the full utilization of ML for clinical data analysis. We investigated the potential of the ChatGPT Advanced Data Analysis (ADA), an extension of GPT-4, to bridge this gap and perform ML analyses efficiently. Real-world clinical datasets and study details from large trials across various medical specialties were presented to ChatGPT ADA without specific guidance. ChatGPT ADA autonomously developed state-of-the-art ML models based on the original study’s training data to predict clinical outcomes such as cancer development, cancer progression, disease complications, or biomarkers such as pathogenic gene sequences. Following the re-implementation and optimization of the published models, the head-to-head comparison of the ChatGPT ADA-crafted ML models and their respective manually crafted counterparts revealed no significant differences in traditional performance metrics (p ≥ 0.072). Strikingly, the ChatGPT ADA-crafted ML models often outperformed their counterparts. In conclusion, ChatGPT ADA offers a promising avenue to democratize ML in medicine by simplifying complex data analyses, yet should enhance, not replace, specialized training and resources, to promote broader applications in medical research and practice.

Abstract Developing robust artificial intelligence (AI) models that generalize well to unseen datasets is challenging and usually requires large and variable datasets, preferably from multiple institutions. In federated learning (FL), a model is trained collaboratively at numerous sites that hold local datasets without exchanging them. So far, the impact of training strategy, i.e., local versus collaborative, on the diagnostic on-domain and off-domain performance of AI models interpreting chest radiographs has not been assessed. Consequently, using 610,000 chest radiographs from five institutions across the globe, we assessed diagnostic performance as a function of training strategy (i.e., local vs. collaborative), network architecture (i.e., convolutional vs. transformer-based), single versus cross-institutional performance (i.e., on-domain vs. off-domain), imaging finding (i.e., cardiomegaly, pleural effusion, pneumonia, atelectasis, consolidation, pneumothorax, and no abnormality), dataset size (i.e., from n = 18,000 to 213,921 radiographs), and dataset diversity. Large datasets not only showed minimal performance gains with FL but, in some instances, even exhibited decreases. In contrast, smaller datasets revealed marked improvements. Thus, on-domain performance was mainly driven by training data size. However, off-domain performance leaned more on training diversity. When trained collaboratively across diverse external institutions, AI models consistently surpassed models trained locally for off-domain tasks, emphasizing FL’s potential in leveraging data diversity. In conclusion, FL can bolster diagnostic privacy, reproducibility, and off-domain reliability of AI models and, potentially, optimize healthcare outcomes.

Abstract Large language models (LLMs) have shown potential in various applications, including clinical practice. However, their accuracy and utility in providing treatment recommendations for orthopedic conditions remain to be investigated. Thus, this pilot study aims to evaluate the validity of treatment recommendations generated by GPT-4 for common knee and shoulder orthopedic conditions using anonymized clinical MRI reports. A retrospective analysis was conducted using 20 anonymized clinical MRI reports, with varying severity and complexity. Treatment recommendations were elicited from GPT-4 and evaluated by two board-certified specialty-trained senior orthopedic surgeons. Their evaluation focused on semiquantitative gradings of accuracy and clinical utility and potential limitations of the LLM-generated recommendations. GPT-4 provided treatment recommendations for 20 patients (mean age, 50 years ± 19 [standard deviation]; 12 men) with acute and chronic knee and shoulder conditions. The LLM produced largely accurate and clinically useful recommendations. However, limited awareness of a patient’s overall situation, a tendency to incorrectly appreciate treatment urgency, and largely schematic and unspecific treatment recommendations were observed and may reduce its clinical usefulness. In conclusion, LLM-based treatment recommendations are largely adequate and not prone to ‘hallucinations’, yet inadequate in particular situations. Critical guidance by healthcare professionals is obligatory, and independent use by patients is discouraged, given the dependency on precise data input.

Abstract Accurate and automatic segmentation of fibroglandular tissue in breast MRI screening is essential for the quantification of breast density and background parenchymal enhancement. In this retrospective study, we developed and evaluated a transformer-based neural network for breast segmentation (TraBS) in multi-institutional MRI data, and compared its performance to the well established convolutional neural network nnUNet. TraBS and nnUNet were trained and tested on 200 internal and 40 external breast MRI examinations using manual segmentations generated by experienced human readers. Segmentation performance was assessed in terms of the Dice score and the average symmetric surface distance. The Dice score for nnUNet was lower than for TraBS on the internal testset (0.909 ± 0.069 versus 0.916 ± 0.067, P

Abstract When clinicians assess the prognosis of patients in intensive care, they take imaging and non-imaging data into account. In contrast, many traditional machine learning models rely on only one of these modalities, limiting their potential in medical applications. This work proposes and evaluates a transformer-based neural network as a novel AI architecture that integrates multimodal patient data, i.e., imaging data (chest radiographs) and non-imaging data (clinical data). We evaluate the performance of our model in a retrospective study with 6,125 patients in intensive care. We show that the combined model (area under the receiver operating characteristic curve [AUROC] of 0.863) is superior to the radiographs-only model (AUROC = 0.811, p

Abstract Although generative adversarial networks (GANs) can produce large datasets, their limited diversity and fidelity have been recently addressed by denoising diffusion probabilistic models, which have demonstrated superiority in natural image synthesis. In this study, we introduce Medfusion, a conditional latent DDPM designed for medical image generation, and evaluate its performance against GANs, which currently represent the state-of-the-art. Medfusion was trained and compared with StyleGAN-3 using fundoscopy images from the AIROGS dataset, radiographs from the CheXpert dataset, and histopathology images from the CRCDX dataset. Based on previous studies, Progressively Growing GAN (ProGAN) and Conditional GAN (cGAN) were used as additional baselines on the CheXpert and CRCDX datasets, respectively. Medfusion exceeded GANs in terms of diversity (recall), achieving better scores of 0.40 compared to 0.19 in the AIROGS dataset, 0.41 compared to 0.02 (cGAN) and 0.24 (StyleGAN-3) in the CRMDX dataset, and 0.32 compared to 0.17 (ProGAN) and 0.08 (StyleGAN-3) in the CheXpert dataset. Furthermore, Medfusion exhibited equal or higher fidelity (precision) across all three datasets. Our study shows that Medfusion constitutes a promising alternative to GAN-based models for generating high-quality medical images, leading to improved diversity and less artifacts in the generated images.

Abstract Recent advances in computer vision have shown promising results in image generation. Diffusion probabilistic models have generated realistic images from textual input, as demonstrated by DALL-E 2, Imagen, and Stable Diffusion. However, their use in medicine, where imaging data typically comprises three-dimensional volumes, has not been systematically evaluated. Synthetic images may play a crucial role in privacy-preserving artificial intelligence and can also be used to augment small datasets. We show that diffusion probabilistic models can synthesize high-quality medical data for magnetic resonance imaging (MRI) and computed tomography (CT). For quantitative evaluation, two radiologists rated the quality of the synthesized images regarding "realistic image appearance", "anatomical correctness", and "consistency between slices". Furthermore, we demonstrate that synthetic images can be used in self-supervised pre-training and improve the performance of breast segmentation models when data is scarce (Dice scores, 0.91 [without synthetic data], 0.95 [with synthetic data]).

Abstract Due to the rapid advancements in recent years, medical image analysis is largely dominated by deep learning (DL). However, building powerful and robust DL models requires training with large multi-party datasets. While multiple stakeholders have provided publicly available datasets, the ways in which these data are labeled vary widely. For Instance, an institution might provide a dataset of chest radiographs containing labels denoting the presence of pneumonia, while another institution might have a focus on determining the presence of metastases in the lung. Training a single AI model utilizing all these data is not feasible with conventional federated learning (FL). This prompts us to propose an extension to the widespread FL process, namely flexible federated learning (FFL) for collaborative training on such data. Using 695,000 chest radiographs from five institutions from across the globe—each with differing labels—we demonstrate that having heterogeneously labeled datasets, FFL-based training leads to significant performance increase compared to conventional FL training, where only the uniformly annotated images are utilized. We believe that our proposed algorithm could accelerate the process of bringing collaborative training methods from research and simulation phase to the real-world applications in healthcare.

Quantitative MRI techniques such as T2 and T1ρ mapping are beneficial in evaluating knee joint pathologies; however, long acquisition times limit their clinical adoption. MIXTURE (Multi-Interleaved X-prepared Turbo Spin-Echo with IntUitive RElaxometry) provides a versatile turbo spin-echo (TSE) platform for simultaneous morphologic and quantitative joint imaging. Two MIXTURE sequences were designed along clinical requirements: “MIX1”, combining proton density (PD)-weighted fat-saturated (FS) images and T2 mapping (acquisition time: 4:59 min), and “MIX2”, combining T1-weighted images and T1ρ mapping (6:38 min). MIXTURE sequences and their reference 2D and 3D TSE counterparts were acquired from ten human cadaveric knee joints at 3.0 T. Contrast, contrast-to-noise ratios, and coefficients of variation were comparatively evaluated using parametric tests. Clinical radiologists (n = 3) assessed diagnostic quality as a function of sequence and anatomic structure using five-point Likert scales and ordinal regression, with a significance level of α = 0.01. MIX1 and MIX2 had at least equal diagnostic quality compared to reference sequences of the same image weighting. Contrast, contrast-to-noise ratios, and coefficients of variation were largely similar for the PD-weighted FS and T1-weighted images. In clinically feasible scan times, MIXTURE sequences yield morphologic, TSE-based images of diagnostic quality and quantitative parameter maps with additional insights on soft tissue composition and ultrastructure.

Abstract Magnetic resonance imaging (MRI) is commonly used to assess traumatic and non-traumatic conditions of the knee. Due to its complex and variable anatomy, the posterolateral corner (PLC)—often referred to as the joint’s dark side—remains diagnostically challenging. We aimed to render the diagnostic evaluation of the PLC more functional by combining MRI, varus loading, and image post-processing in a model of graded PLC injury that used sequential transections of the lateral collateral ligament, popliteus tendon, popliteofibular ligament, and anterior cruciate ligament. Ten human cadaveric knee joint specimens underwent imaging in each condition as above, and both unloaded and loaded using an MR-compatible device that standardized loading (of 147 N) and position (at 30° flexion). Following manual segmentation, 3D joint models were used to computationally measure lateral joint space opening for each specimen, configuration, and condition, while manual measurements provided the reference standard. With more extensive ligament deficiency and loading, lateral joint spaces increased significantly. In conclusion, varus stress MRI allows comprehensive PLC evaluation concerning structural integrity and associated functional capacity. Beyond providing normative values of lateral compartment opening, this study has potential implications for diagnostic and surgical decision-making and treatment monitoring in PLC injuries.

Increased attention has been given to MRI in radiation-free screening for malignant nodules in recent years. Our objective was to compare the performance of human readers and radiomic feature analysis based on stand-alone and complementary CT and MRI imaging in classifying pulmonary nodules. This single-center study comprises patients with CT findings of pulmonary nodules who underwent additional lung MRI and whose nodules were classified as benign/malignant by resection. For radiomic features analysis, 2D segmentation was performed for each lung nodule on axial CT, T2-weighted (T2w), and diffusion (DWI) images. The 105 extracted features were reduced by iterative backward selection. The performance of radiomics and human readers was compared by calculating accuracy with Clopper–Pearson confidence intervals. Fifty patients (mean age 63 +/− 10 years) with 66 pulmonary nodules (40 malignant) were evaluated. ACC values for radiomic features analysis vs. radiologists based on CT alone (0.68; 95%CI: 0.56, 0.79 vs. 0.59; 95%CI: 0.46, 0.71), T2w alone (0.65; 95%CI: 0.52, 0.77 vs. 0.68; 95%CI: 0.54, 0.78), DWI alone (0.61; 95%CI:0.48, 0.72 vs. 0.73; 95%CI: 0.60, 0.83), combined T2w/DWI (0.73; 95%CI: 0.60, 0.83 vs. 0.70; 95%CI: 0.57, 0.80), and combined CT/T2w/DWI (0.83; 95%CI: 0.72, 0.91 vs. 0.64; 95%CI: 0.51, 0.75) were calculated. This study is the first to show that by combining quantitative image information from CT, T2w, and DWI datasets, pulmonary nodule assessment through radiomics analysis is superior to using one modality alone, even exceeding human readers’ performance.

Based on in silico, in vitro, in situ, and in vivo evaluations, this study aims to establish and optimize the chemical exchange saturation transfer (CEST) imaging of lactate (Lactate-CEST—LATEST). To this end, we optimized LATEST sequences using Bloch–McConnell simulations for optimal detection of lactate with a clinical 3 T MRI scanner. The optimized sequences were used to image variable lactate concentrations in vitro (using phantom measurements), in situ (using nine human cadaveric lower leg specimens), and in vivo (using four healthy volunteers after exertional exercise) that were then statistically analyzed using the non-parametric Friedman test and Kendall Tau-b rank correlation. Within the simulated Bloch–McConnell equations framework, the magnetization transfer ratio asymmetry (MTRasym) value was quantified as 0.4% in the lactate-specific range of 0.5–1 ppm, both in vitro and in situ, and served as the imaging surrogate of the lactate level. In situ, significant differences (p < 0.001) and strong correlations (τ = 0.67) were observed between the MTRasym values and standardized intra-muscular lactate concentrations. In vivo, a temporary increase in the MTRasym values was detected after exertional exercise. In this bench-to-bedside comprehensive feasibility study, different lactate concentrations were detected using an optimized LATEST imaging protocol in vitro, in situ, and in vivo at 3 T, which prospectively paves the way towards non-invasive quantification and monitoring of lactate levels across a broad spectrum of diseases.

Abstract Abnormal torsion of the lower limbs may adversely affect joint health. This study developed and validated a deep learning-based method for automatic measurement of femoral and tibial torsion on MRI. Axial T2-weighted sequences acquired of the hips, knees, and ankles of 93 patients (mean age, 13 ± 5 years; 52 males) were included and allocated to training (n = 60), validation (n = 9), and test sets (n = 24). A U-net convolutional neural network was trained to segment both femur and tibia, identify osseous anatomic landmarks, define pertinent reference lines, and quantify femoral and tibial torsion. Manual measurements by two radiologists provided the reference standard. Inter-reader comparisons were performed using repeated-measures ANOVA, Pearson’s r, and the intraclass correlation coefficient (ICC). Mean Sørensen-Dice coefficients for segmentation accuracy ranged between 0.89 and 0.93 and erroneous segmentations were scarce. Ranges of torsion as measured by both readers and the algorithm on the same axial image were 15.8°–18.0° (femur) and 33.9°–35.2° (tibia). Correlation coefficients (ranges, .968 ≤ r ≤ .984 [femur]; .867 ≤ r ≤ .904 [tibia]) and ICCs (ranges, .963 ≤ ICC ≤ .974 [femur]; .867 ≤ ICC ≤ .894 [tibia]) indicated excellent inter-reader agreement. Algorithm-based analysis was faster than manual analysis (7 vs 207 vs 230 s, p

Abstract Clinical Magnetic Resonance Imaging (MRI) of joints is limited to mere morphologic evaluation and fails to directly visualize joint or ligament function. In this controlled laboratory study, we show that knee joint functionality may be quantified in situ and as a function of graded posterior cruciate ligament (PCL)-deficiency by combining MRI and standardized loading. 11 human knee joints underwent MRI under standardized posterior loading in the unloaded and loaded (147 N) configurations and in the intact, partially, and completely PCL-injured conditions. For each specimen, configuration, and condition, 3D joint models were implemented to analyse joint kinematics based on 3D Euclidean vectors and their projections on the Cartesian planes. Manual 2D measurements served as reference. With increasing PCL deficiency, vector projections increased significantly in the anteroposterior dimension under loading and manual measurements demonstrated similar patterns of change. Consequently, if combined with advanced image post-processing, stress MRI is a powerful diagnostic adjunct to evaluate ligament functionality and joint laxity in multiple dimensions and may have a role in differentiating PCL injury patterns, therapeutic decision-making, and treatment monitoring.

Abstract For proximal femoral nailing, choosing the proper entry point with the aid of C-arm imaging is crucial. Therefore, obtaining accurate radiological views that facilitate sound identification of the tip of the greater trochanter (GT) is of utmost importance. The aim of this study was to define a radiological view characterised by reproducible radiographic landmarks which will allow the reliable identification of the tip of the GT in the anteroposterior view. Anatomical and radiographic features of 16 cadaveric femurs were analysed. The cortical overlap view (COV), characterised by the radiological overlap of the density line of the piriform fossa and the intertrochanteric crest, was identified. It marks the rotation of the proximal femur at which the GT can be accurately identified and used to determine the desired entry point for a proximal femoral nail. Trainees and fully qualified orthopedic trauma surgeons were asked to identify the correct COV in radiological imaging series. Mean internal rotation of the femur to achieve a COV was 17.5° (range 12.8°–21.8°). In the COV the tip of the GT is the highest visible point and the mean distance from the cortical overlap line to the tip of the GT is 4.45 mm. Intra- and inter-rater reliability was high with ICC(2,k) = 0.932 and ICC(2,k) = 0.987 respectively. Trainees achieved higher rates of correct COV identification than specialists. There was no significant correlation between the internal rotation of the femur to achieve the COV and femoral antetorsion. In conclusion, the COV is a highly reproducible radiological view that is characterised by radiographic landmarks easy to recognise. It allows for accurate identification of the tip of the GT, which can be used by the surgeon as a reference to determine the desired entry point for an intramedullary nail.

Unmasking the decision making process of machine learning models is essential for implementing diagnostic support systems in clinical practice. Here, the authors demonstrate that adversarially trained models can significantly enhance the usability of pathology detection as compared to their standard counterparts.

Abstract Background To compare the glycosaminoglycan (GAG) content of lumbar intervertebral disks (IVDs) of patients with ankylosing spondylitis (AS) and healthy volunteers and to investigate the association of GAG depletion and disease-related clinical and imaging features. Methods Lumbar spines of 50 AS patients (mean age 50 ± 10.5 years) and 30 age-matched volunteers were studied with 3-T magnetic resonance imaging (MRI) and conventional radiographs (CR). The MRI protocol included high-resolution morphological sequences and the compositional GAG chemical exchange saturation transfer imaging technique (gagCEST). Morphological images were analyzed by three raters for inflammatory activity, fat deposition, disk degeneration, and structural changes on CR. Clinical and serological measures included the Bath AS Disease Activity (BASDAI) and Bath AS Function (BASFI) Indices and C-reactive protein (CRP) levels. GagCEST values of both groups were compared using a linear mixed model. Kendall-Tau correlation analyses were performed. Results GagCEST values were significantly lower in AS patients (2.0 ± 1.7%) vs. healthy volunteers (2.4 ± 1.8%), p = 0.001. Small, yet significant correlations were found between gagCEST values and CRP levels (τ = − 0.14, p = 0.007), BASFI (τ = − 0.18, p

Abstract Background Even though cartilage loss is a known feature of psoriatic arthritis (PsA), little is known about its role in the pathogenesis of PsA. Using delayed gadolinium-enhanced magnetic resonance imaging of cartilage (dGEMRIC) as a non-invasive marker of the tissue’s proteoglycan content, such early (i.e., pre-morphological) changes have been associated with inflammation in rheumatoid arthritis (RA). Yet, this association has not been studied before in PsA. Methods The metacarpophalangeal (MCP), proximal interphalangeal (PIP), and distal interphalangeal (DIP) joints of 17 patients with active PsA were evaluated by high-resolution clinical standard morphological and dGEMRIC sequences using a 3T MRI scanner (Magnetom Skyra, Siemens) and a dedicated 16-channel hand coil. Images were analyzed by two independent raters for dGEMRIC indices, PsA MRI scores (PsAMRIS), and total cartilage thickness (TCT). Kendall tau correlation coefficients (τ) were calculated. Results We found significant negative correlations between dGEMRIC indices and total PsAMRIS (τ = − 0.5, p = 0.012), synovitis (τ = − 0.56, p = 0.006), flexor tenosynovitis (τ = − 0.4, p = 0.049), and periarticular inflammation (τ = − 0.72, p

Abstract Background To evaluate the value of 3 Tesla (T) magnetic resonance imaging (MRI) changes of flexor tendon pulleys for the differentiation of psoriatic (PsA) and rheumatoid arthritis (RA), using a novel 16-channel high-resolution hand coil. Methods Seventeen patients with active PsA, 20 patients with active RA, and 16 healthy controls (HC) underwent high-resolution 3 T MRI using the dedicated 16-channel hand coil. Images were analyzed by three independent readers for the degree of inflammatory changes, thickness of flexor tendon pulleys, and comparison to the outcome measures for RA clinical trials (OMERACT) PsA MRI score (PsAMRIS) and to its sub-scores. For correlation analyses, Spearman rho correlation was calculated. Results Flexor tendon pulleys were thicker in PsA than in RA patients (mean difference 0.16 mm, p