Ab0931 Validation of the Computer Analysis of Ultrasound Images in the Study of Tendons and Entheses of Interest in Spondyloarthropathies

Background: In patients with spondyloarthritis, ultrasound assessment of tendons and entheses is performed using ordinal scales based on findings that each operator interprets based on his knowledge and experience. Despite this, inter-observer and intra-observer reliability has been shown to be good or very good. However, we do not have properly validated quantitative methods of echo-structural assessment of entheses and tendons. Objectives: To validate the use of ImageJ software in the study of tendons and entheses in healthy subjects and those diagnosed with spondyloarthritis in different longitudinal scanning planes. Methods: Ten healthy volunteers aged 35-50 years (5 women and 5 men), with no musculoskeletal symptoms and no family history of spondyloarthritis; and 10 patients with spondyloarthritis (5 with axial spondyloarthritis and 5 with psoriatic arthritis) of the same age group and sex as controls were recruited. A single rheumatologist with expertise in ultrasound performed a standard ultrasound study following the MASEI protocol for the study of entheses, with a GE S8 ultrasound scanner equipped with an 8 to 15MHz linear probe (Gn 55, dynamic range 66 dB). All scans included two longitudinal evaluations by modifying the angle of incidence between 30 and 90º (Figure 1A) except for the plantar fascia. Two different rheumatologists used the ImageJ software v1.53e (Wayne Rasband, NIH, USA) (Figure 1B) The agreement between the mean and mode of the greyscale determinations and the standard deviation of the mean of the two determinations made was determined and expressed in terms of percentage error between the highest and lowest measurement. Results: In general, the interobserver and intra-observer agreement of the more superficial tendons (patellar, Achilles) was higher than that of the deep tendons (plantar fascia, quadriceps). Inter-observer and intra-observer agreement in vertical planes was higher than in non-vertical planes, especially in non-tubular tendons (quadriceps, triceps and plantar fascia). In tubular tendons the inter-plane agreement was higher than in flat tendons (Table 1). The same analyses performed among patients exclusively showed lower interobserver agreement in those tendons with the presence of enthesophytosis in both planes (19.6% difference between measurements) compared to that of patients without enthesophytes (9.8%). The same effect was found in patients with intratendinous calcifications. Conclusion: Computer analysis of static images is a tool that demonstrates very good inter-observer and intra-observer agreement in most of the tendons included in the MASEI study, especially those with tubular conformation. As a limitation, it has been observed that the stigmata of spondyloarthritis chronicity alter the measurements performed and worsen the reliability of the results. REFERENCES: NIL. Acknowledgements: NIL. Disclosure of Interests: None declared.