Search

In recent years, much research evaluating the radiographic destruction of finger joints in patients with rheumatoid arthritis (RA) using deep learning models was conducted. Unfortunately, most previous models were not clinically applicable due to the small object regions as well as the close spatial relationship. In recent years, a new network structure called RetinaNets, in combination with the focal loss function, proved reliable for detecting even small objects. Therefore, the study aimed to increase the recognition performance to a clinically valuable level by proposing an innovative approach with adaptive changes in intersection over union (IoU) values during training of Retina Networks using the focal loss error function. To this end, the erosion score was determined using the Sharp van der Heijde (SvH) metric on 300 conventional radiographs from 119 patients with RA. Subsequently, a standard RetinaNet with different IoU values as well as adaptively modified IoU values were trained and compared in terms of accuracy, mean average accuracy (mAP), and IoU. With the proposed approach of adaptive IoU values during training, erosion detection accuracy could be improved to 94% and an mAP of 0.81 ± 0.18. In contrast Retina networks with static IoU values achieved only an accuracy of 80% and an mAP of 0.43 ± 0.24. Thus, adaptive adjustment of IoU values during training is a simple and effective method to increase the recognition accuracy of small objects such as finger and wrist joints.

Currently, clinical indications for the application of gadolinium-based contrast agents (GBCA) in magnetic resonance imaging (MRI) are increasingly being questioned. Consequently, this study aimed to evaluate the additional diagnostic value of contrast enhancement in MRI of the hand in patients with rheumatoid arthritis (RA). Thirty-one patients with RA (mean age, 50 ± 14 years (range, 18–72 years)) underwent morphologic MRI scans on a clinical 3 T scanner. MRI studies were analyzed based on (1) the Rheumatoid Arthritis Magnetic Resonance Imaging Score (RAMRIS) and (2) the GBCA-free RAMRIS version, termed RAMRIS Sine-Gadolinium-For-Experts (RAMRIS-SAFE), in which synovitis and tenosynovitis were assessed using the short-tau inversion-recovery sequence instead of the post-contrast T1-weighted sequence. The synovitis subscores in terms of Spearman’s ρ, as based on RAMRIS and RAMRIS-SAFE, were almost perfect (ρ = 0.937; p < 0.001), while the tenosynovitis subscores were less strongly correlated (ρ = 0.380 p = 0.035). Correlation between the total RAMRIS and RAMRIS-SAFE was also almost perfect (ρ = 0.976; p < 0.001). Inter-rater reliability in terms of Cohen’s κ was high (0.963 ≤ κ ≤ 0.925). In conclusion, RAMRIS-SAFE as the GBCA-free version of the well-established RAMRIS is a patient-friendly and resource-efficient alternative for assessing disease-related joint changes in RA. As patients with RA are subject to repetitive GBCA applications, non-contrast imaging protocols should be considered.

Background: Even though cartilage loss is a known feature of psoriatic (PsA) and rheumatoid arthritis (RA), research is sparse on its role in the pathogenesis of PsA, its potential use for disease monitoring and for differentiation from RA. We therefore assessed the use of delayed gadolinium-enhanced magnetic resonance imaging of cartilage (dGEMRIC) to evaluate biochemical cartilage changes in metacarpophalangeal (MCP) and proximal interphalangeal (PIP) joints in PsA patients and compared these to RA patients. Materials and Methods: A total of 17 patients with active PsA and 20 patients with active RA were evaluated by high-resolution 3 Tesla dGEMRIC using a dedicated 16-channel hand coil. Images were analyzed by two independent raters for dGEMRIC indices and joint space width (JSW) at MCP and PIP joint levels. Results: No significant differences of dGEMRIC values could be found between both study populations (PsA 472.25 ms, RA 461.11 ms; p = 0.763). In all RA and most PsA patients, PIP joints showed significantly lower dGEMRIC indices than MCP joints (RA: D2: p = 0.009, D3: p = 0.008, D4: p = 0.002, D5: p = 0.002; PsA: D3: p = 0.001, D4: p = 0.004). Most joint spaces had similar widths in both disease entities and no significant differences were found. Conclusions: As evaluated by dGEMRIC, the molecular composition of the MCP and PIP joint cartilage of PsA patients is similar to that of RA patients, demonstrating the scientific and clinical feasibility of compositional magnetic resonance (MR) imaging in these disease entities. Patterns and severity of compositional cartilage degradation of the finger joints may therefore be assessed beyond mere morphology in PsA and RA patients.

Background: To evaluate whether a simplified (s) version of the psoriatic arthritis magnetic resonance imaging score (PsAMRIS), sPsAMRIS, is a potential tool for therapy monitoring in psoriatic arthritis (PsA). Methods: Seventeen patients with active psoriatic arthritis (PsA) underwent magnetic resonance imaging (MRI) at 3 T of the clinically dominant hand at baseline and after 6 months. Scoring was performed by two musculoskeletal radiologists in terms of the PsAMRIS and sPsAMRIS, which is a simplified version with reduced item numbers based on prior evaluation of responsiveness to change by standardized response means (SRMs). Both scores were compared by calculation of overall and each sub-score’s SRMs and relative efficacy (RE) after bootstrapping. Results: PsAMRIS sub-scores of MCP joints 3 and 4, and proximal interphalangeal (PIP) joint 4 had the highest SRM (−0.07 each), indicating highest responsiveness to change, and were, therefore, included in sPsAMRIS. Compared to PsAMRIS, sPsAMRIS was characterized by higher SRMs (sPsAMRIS: −0.13 vs. PsAMRIS: −0.02) and higher RE (29.46). sPsAMRIS and PsAMRIS were highly correlated at baseline (r = 0.75, p < 0.01 (Pearson’s correlation)) and at 6-month follow-up (r = 0.64, p = 0.01). Mean time burden for completion of scoring per MRI study was significantly reduced when using PsAMRIS (469 ± 87.03 s) as compared to sPsAMRIS (140.1 ± 21.25 s) (p < 0.001). Conclusion: Due to its similar responsiveness to change compared to standard PsAMRIS, and time efficiency, sPsAMRIS might be a potential diagnostic tool to quantitatively assess and monitor therapy in PsA.