Inflammation of the Sacroiliac Joints and Spine and Structural Changes on Magnetic Resonance Imaging in Axial Spondyloarthritis: Five-year Data From the DESIR Cohort

Implementation of artificial intelligence models in magnetic resonance imaging with focus on diagnosis of rheumatoid arthritis and axial spondyloarthritis: narrative review

Early diagnosis in rheumatoid arthritis (RA) and axial spondyloarthritis (axSpA) is essential to initiate timely interventions, such as medication and lifestyle changes, preventing irreversible joint damage, reducing symptoms, and improving long-term outcomes for patients. Since magnetic resonance imaging (MRI) of the wrist and hand, in case of RA and MRI of the sacroiliac joints (SIJ) in case of axSpA can identify inflammation before it is clinically discernible, this modality may be crucial for early diagnosis. Artificial intelligence (AI) techniques, together with machine learning (ML) and deep learning (DL) have quickly evolved in the medical field, having an important role in improving diagnosis, prognosis, in evaluating the effectiveness of treatment and monitoring the activity of rheumatic diseases through MRI. The improvements of AI techniques in the last years regarding imaging interpretation have demonstrated that a computer-based analysis can equal and even exceed the human eye. The studies in the field of AI have investigated how specific algorithms could distinguish between tissues, diagnose rheumatic pathology and grade different signs of early inflammation, all of them being crucial for tracking disease activity. The aim of this paper is to highlight the implementation of AI models in MRI with focus on diagnosis of RA and axSpA through a literature review.

Researchers in rheumatology should avoid categorization of continuous predictor variables

BACKGROUND

Rheumatology researchers often categorize continuous predictor variables. We aimed to show how this practice may alter results from observational studies in rheumatology.

METHODS

We conducted and compared the results of two analyses of the association between our predictor variable (percentage change in body mass index [BMI] from baseline to four years) and two outcome variable domains of structure and pain in knee and hip osteoarthritis. These two outcome variable domains covered 26 different outcomes for knee and hip combined. In the first analysis (categorical analysis), percentage change in BMI was categorized as ≥ 5% decrease in BMI, < 5% change in BMI, and ≥ 5% increase in BMI, while in the second analysis (continuous analysis), it was left as a continuous variable. In both analyses (categorical and continuous), we used generalized estimating equations with a logistic link function to investigate the association between the percentage change in BMI and the outcomes.

RESULTS

For eight of the 26 investigated outcomes (31%), the results from the categorical analyses were different from the results from the continuous analyses. These differences were of three types: 1) for six of these eight outcomes, while the continuous analyses revealed associations in both directions (i.e., a decrease in BMI had one effect, while an increase in BMI had the opposite effect), the categorical analyses showed associations only in one direction of BMI change, not both; 2) for another one of these eight outcomes, the categorical analyses suggested an association with change in BMI, while this association was not shown in the continuous analyses (this is potentially a false positive association); 3) for the last of the eight outcomes, the continuous analyses suggested an association of change in BMI, while this association was not shown in the categorical analyses (this is potentially a false negative association).

CONCLUSIONS

Categorization of continuous predictor variables alters the results of analyses and could lead to different conclusions; therefore, researchers in rheumatology should avoid it.

Can radiomics replace the SPARCC scoring system in evaluating bone marrow edema of sacroiliac joints in patients with axial spondyloarthritis?

OBJECTIVES

To develop an objective and efficient method based on radiomics to evaluate bone marrow edema (BMO) of sacroiliac joints (SIJs) by magnetic resonance imaging (MRI) in patients with axial spondyloarthritis (axSpA) and to compare with the Spondyloarthritis Research Consortium of Canada (SPARCC) scoring system.

METHODS

From September 2013 to March 2022, patients with axSpA who underwent 3.0T SIJ-MRI were included and were randomly divided into training and validation cohorts at a ratio of 7:3. The optimal radiomics features selected from the SIJ-MRI in the training cohort were included to generate the radiomics model. The performance of the model was evaluated by ROC analysis and decision curve analysis (DCA). Rad scores were calculated using the radiomics model. The responsiveness was compared for Rad scores and SPARCC scores. We also assessed the correlation between the Rad score and SPARCC score.

RESULTS

A total of 558 patients were finally included. The radiomics model showed favorable discrimination of a SPARCC score <2 or ≥2 both in the training (AUC, 0.90; 95% CI: 0.87-0.93) and validation cohorts (AUC, 0.90; 95% CI, 0.86-0.95). DCA confirmed that the model was clinically useful. Rad score showed higher responsiveness to treatment-related change than SPARCC score. Furthermore, a significant correlation was noted between the Rad score and SPARCC score when scoring the status of BMO (r_s=0.80, P < 0.001), and a strong correlation was noted when scoring the change in BMO (r=0.70, P < 0.001).

CONCLUSION

The study proposed a radiomics model to accurately quantify the BMO of SIJs in patients with axSpA, providing an alternative to the SPARCC scoring system. Key Points • The Rad score is an index with high validity for the objective and quantitative evaluation of bone marrow edema (BMO) of the sacroiliac joints in axial spondyloarthritis. • The Rad score is a promising tool to monitor the change of BMO upon treatment.

MRI in axial spondyloarthritis: understanding an 'ASAS-positive MRI' and the ASAS classification criteria

In 2009, the Assessment of SpondyloArthritis international Society (ASAS) published a definition of 'active sacroiliitis on magnetic resonance imaging (MRI) for classification of axial spondyloarthritis'. This new definition of an 'ASAS-positive MRI' was integral to new classification criteria for axial spondyloarthritis that were published in the same year. The ASAS MRI definition had the considerable advantage of simplicity and the definition gained popularity as guidance for interpreting MRI of the sacroiliac joints in clinical practice. However, classification criteria are not designed for use in clinical practice with the consequence that overreliance on the presence of bone marrow edema, which is the principal determinant of an 'ASAS-positive MRI', may result in a tendency to overcall inflammatory sacroiliitis in the clinical setting. This article aims to inform the reader about the rationale behind the ASAS definition of a positive MRI and ASAS classification criteria, their proper use in research and why they should not be used in clinical practice. The article also contains guidance for an updated imaging protocol and interpretation of images including typical imaging findings, differential diagnosis, and common pitfalls.