Correction: Biomarkers of erosive arthritis in systemic lupus erythematosus: Application of machine learning models

Musculoskeletal involvement in systemic lupus erythematosus: a contrast-enhanced magnetic resonance imaging study in 107 subjects

OBJECTIVE

Joint involvement in SLE is the most frequent manifestation and shows a wide heterogeneity. It has not a valid classification and it is often underestimated. Subclinical inflammatory musculoskeletal involvement is not well known. We aim to describe the prevalence of joint and tendon involvement in hand and wrist of SLE patients, either with clinical arthritis, arthralgia or asymptomatic and compare it with healthy subjects using contrasted MRI.

METHODS

SLE patients fulfilling SLICC criteria were recruited and classified as follows: group (G) 1: hand/wrist arthritis, G2: hand/wrist arthralgia, G3: no hand/wrist symptoms. Jaccoud arthropathy, CCPa and RF positivity, hand OA or surgery were excluded. Healthy subjects (HS) were recruited as controls: G4. Contrasted MRI of non-dominant hand/wrist was performed. Images were evaluated following RAMRIS criteria extended to PIP, Tenosynovitis score for RA and peritendonitis from PsAMRIS. Groups were statistically compared.

RESULTS

A total of 107 subjects were recruited (G1: 31, G2:31, G3:21, G4:24). Any lesion: SLE patients 74.7%, HS 41.67%; P 0.002. Synovitis: G1: 64.52%, G2: 51.61%, G3: 45%, G4: 20.83%; P 0.013. Erosions: G1: 29.03%; G2: 54.84%, G3: 47.62%; G4: 25%; P 0.066. Bone marrow oedema: G1: 29.03%, G2: 22.58%, G3: 19.05%, G4: 0.0%; P 0.046. Tenosynovitis: G1: 38.71%; G2: 25.81%, G3: 14.29%, G4: 0.0%; P 0.005. Peritendonitis: G1: 12.90%; G2: 3.23%, G3: 0.0%, G4: 0.0%; P 0.07.

CONCLUSION

SLE patients have a high prevalence of inflammatory musculoskeletal alterations confirmed by contrasted MRI, even if asymptomatic. Not only tenosynovitis but peritendonitis is also present.

A pilot study of a machine-learning tool to assist in the diagnosis of hand arthritis

OBJECTIVE

Arthritis is a common condition, which frequently involves the hands. Patients with inflammatory arthritis have been shown to experience significant delays in diagnosis. We sought to develop and test a screening tool combining an image of a patient's hands, a short series of questions, and a single examination technique, to determine the most likely diagnosis in a patient presenting with hand arthritis. Machine learning techniques were used to develop separate algorithms for each component, which were combined to produce a diagnosis.

METHODS

280 consecutive new patients presenting to a Rheumatology practice with hand arthritis were enrolled. Each patient completed a 9-part questionnaire, had photographs taken of each hand, and had a single examination result recorded. The Rheumatologist diagnosis was recorded following a 45-minute consultation. The photograph algorithm was developed from 1000 previous hand images, and machine learning techniques were applied to the questionnaire results, training several models against the diagnosis from the Rheumatologist.

RESULTS

The combined algorithms in this study were able to predict inflammatory arthritis with an accuracy, precision, recall and specificity of 96·8%, 97·2%, 98·6% and 90·5% respectively. Similar results were found when inflammatory arthritis was subclassified into rheumatoid arthritis and psoriatic arthritis. The corresponding figures for osteoarthritis were 79·6%, 85·9%, 61·9% and 92·6%.

CONCLUSION

This study demonstrates a novel application combining image-processing and a patient questionnaire with applied machine-learning methods, to facilitate the diagnosis of patients presenting with hand arthritis. Preliminary results are encouraging for the application of such techniques in clinical practice. This article is protected by copyright. All rights reserved.