Prediction of the activity of early ankylosing spondylitis using radiomics texture analysis on short tau inversion recovery (STIR)

The study aimed to explore the value of texture analysis of radiomics based on the short tau inversion recovery (STIR) sequence to evaluate the activity of bone marrow oedema of sacroiliac joints in early AS.

43 patients with early AS whose data were randomly divided into the training cohort (n=116) and verification cohort (n=56) according to the ratio of 7:3. The optimal feature subsets were obtained by Mann-Whitney U-test, the minimum-Redundancy Maximum-Relevancy (mRMR), and then least absolute shrinkage and selection operator (LASSO) using these texture feature parameters, which were used to construct the final prediction model and obtained the Radscore. The ROC curve was performed to evaluate the performance of the model. The Spearman correlation test was used to analyse the correlation of various indicators.

In the training cohort, to differentiate early AS sacroiliac joint bone marrow oedema between the active and stable groups, the AUCs of the Radscore, SPARCC and ADC were 0.81, 0.91, 0.78, respectively. In the validation cohort, the AUCs were 0.87, 0.89, 0.85. In the two cohorts, there were no significant differences in AUCs between values of the Radscore and SPARCC, ADC (p>0.05). There was a significant difference in AUC between SPARCC and ADC in the training cohort (p<0.05), with no statistical significance in the validation cohort (p>0.05). The correlations were all low between the Radscore values and the values of ESR, CRP, tI, ASDAS-ESR and ASDAS-CRP (p<0.05).

Radiomics analysis based on STIR texture analysis has a good prediction for the evaluation of bone marrow oedema activity of sacroiliac joints in AS. It can be a new non-invasive and objective evaluation method for AS activity.