Applying cascaded convolutional neural network design further enhances automatic scoring of arthritis disease activity on ultrasound images from rheumatoid arthritis patients

• Published in: Annals of the rheumatic diseases Vol. 79; no. 9; pp. 1189 - 1193
• Main Authors: Christensen, Anders Bossel Holst, Just, Søren Andreas, Andersen, Jakob Kristian Holm, Savarimuthu, Thiusius Rajeeth
• Format: Journal Article
• Language: English
• Published: England BMJ Publishing Group Ltd and European League Against Rheumatism 01.09.2020; Elsevier Limited
• Subjects: Accuracy; Adult; Algorithms; Application programming interface; Arthritis, Rheumatoid - diagnostic imaging; Biopsy; Classification; Clinical medicine; Clinical Trials as Topic; Datasets; Design; disease activity; Female; Humans; Image Interpretation, Computer-Assisted - methods; Joints - diagnostic imaging; Male; Middle Aged; Neural networks; Neural Networks, Computer; Patients; Reproducibility of Results; Rheumatoid arthritis; Rheumatology - methods; Severity of Illness Index; Synovitis; Synovitis - diagnostic imaging; Ultrasonic imaging; ultrasonography; Ultrasonography - methods; Ultrasonography - statistics & numerical data; Ultrasound; disease activity; ultrasonography; rheumatoid arthritis
• ISSN: 0003-4967; 1468-2060; 1468-2060
• DOI: 10.1136/annrheumdis-2019-216636

ObjectivesWe have previously shown that neural network technology can be used for scoring arthritis disease activity in ultrasound images from rheumatoid arthritis (RA) patients, giving scores according to the EULAR-OMERACT grading system. We have now further developed the architecture of this neural network and can here present a new idea applying cascaded convolutional neural network (CNN) design with even better results. We evaluate the generalisability of this method on unseen data, comparing the CNN with an expert rheumatologist.MethodsThe images were graded by an expert rheumatologist according to the EULAR-OMERACT synovitis scoring system. CNNs were systematically trained to find the best configuration. The algorithms were evaluated on a separate test data set and compared with the gradings of an expert rheumatologist on a per-joint basis using a Kappa statistic, and on a per-patient basis using a Wilcoxon signed-rank test.ResultsWith 1678 images available for training and 322 images for testing the model, it achieved an overall four-class accuracy of 83.9%. On a per-patient level, there was no significant difference between the classifications of the model and of a human expert (p=0.85). Our original CNN had a four-class accuracy of 75.0%.ConclusionsUsing a new network architecture we have further enhanced the algorithm and have shown strong agreement with an expert rheumatologist on a per-joint basis and on a per-patient basis. This emphasises the potential of using CNNs with this architecture as a strong assistive tool for the objective assessment of disease activity of RA patients.