Search

Objectives The first objective of this study was to implement and assess the performance and reliability of a vision transformer (ViT)-based deep-learning model, an ‘off-the-shelf’ artificial intelligence solution, for identifying distinct signs of microangiopathy in nailfold capilloroscopy (NFC) images of patients with SSc. The second objective was to compare the ViT’s analysis performance with that of practising rheumatologists. Methods NFC images of patients prospectively enrolled in our European Scleroderma Trials and Research group (EUSTAR) and Very Early Diagnosis of Systemic Sclerosis (VEDOSS) local registries were used. The primary outcome investigated was the ViT’s classification performance for identifying disease-associated changes (enlarged capillaries, giant capillaries, capillary loss, microhaemorrhages) and the presence of the scleroderma pattern in these images using a cross-fold validation setting. The secondary outcome involved a comparison of the ViT’s performance vs that of rheumatologists on a reliability set, consisting of a subset of 464 NFC images with majority vote–derived ground-truth labels. Results We analysed 17 126 NFC images derived from 234 EUSTAR and 55 VEDOSS patients. The ViT had good performance in identifying the various microangiopathic changes in capillaries by NFC [area under the curve (AUC) from 81.8% to 84.5%]. In the reliability set, the rheumatologists reached a higher average accuracy, as well as a better trade-off between sensitivity and specificity compared with the ViT. However, the annotators’ performance was variable, and one out of four rheumatologists showed equal or lower classification measures compared with the ViT. Conclusions The ViT is a modern, well-performing and readily available tool for assessing patterns of microangiopathy on NFC images, and it may assist rheumatologists in generating consistent and high-quality NFC reports; however, the final diagnosis of a scleroderma pattern in any individual case needs the judgement of an experienced observer.

Additional file 1.

ObjectivesTo identify the predictive clinical characteristics and establish a prediction model for the progression of mild interstitial lung disease (ILD) in patients with systemic sclerosis (SSc).MethodsPatients with SSc from two independent prospective cohorts were included in this observational study. All patients fulfilled the 2013 American College of Rheumatology/European League Against Rheumatism criteria, had mild ILD at baseline diagnosed by High-Resolution Computed Tomography (HRCT), available baseline and ≥1 annual follow-up pulmonary function tests and no concomitant pulmonary hypertension or airflow obstruction. ILD progression was defined as a relative decrease in forced vital capacity (FVC)%≥15%, or FVC%≥10% combined with diffusing capacity for carbon monoxide %≥15% at 1-year follow-up. Candidate predictors for multivariate logistic regression were selected by expert opinion based on clinical significance. A prediction model for ILD progression was established in the derivation cohort and validated in the multinational validation cohort.ResultsA total of 25/98 and 25/117 patients with SSc showed ILD progression in the derivation cohort and the validation cohort, respectively. Lower SpO2 after 6 min walk test (6MWT) and arthritis ever were identified as independent predictors for ILD progression in both cohorts. The optimal cut-off value of SpO 2 after 6MWT for predicting ILD progression was determined as 94% by receiver operating characteristic curve analysis. The derived SPAR model combining both predictors (SPO2 and ARthritis) increased the prediction rate from 25.5% to 91.7% with an area under the curve (95% CI) of 0.83 (0.73 to 0.93).ConclusionsThe evidence-based SPAR prediction model developed in our study might be helpful for the risk stratification of patients with mild SSc-ILD in clinical practice and cohort enrichment for future clinical trial design. This article has been accepted for publication in 'Annals of the Rheumatic Diseases', 2018 following peer review, and the Version of Record can be accessed online at https://doi.org/10.1136/annrheumdis-2018-213201. Deposited by shareyourpaper.org and openaccessbutton.org. We've taken reasonable steps to ensure this content doesn't violate copyright. However, if you think it does you can request a takedown by emailing help@openaccessbutton.org.