Computer Vision News - February 2020

2 Summary Medical Imaging Challenge 4 Individuals with Rheumatoid Arthritis (RA) suffer from pain and sti ffness in the joints. Damage in affected joints is due to inf lammation of the synovium. By assessing hands and feet erosion, radiography may reveal the status of the joints. Assessment of damage is key to measure the severity of the disease, as wel l as to monitor patient response to treatment. The measurement of damage is currently done via a manual inspection of radiographic images to score bony erosion and joint space narrowing . Scoring of radiographs is general ly time- consuming and subjective. An automated approach is highly desirable, since it wi l l provide an objective evaluation of joint damage, free of inter- and intra- variation. The most commonly used approach to grade radiographs in RA is the Sharp/van der Hei jde (SvH) scoring method for joint erosion. The SvH method includes: in each hand, 16 areas for erosions and 15 areas for joint space narrowing; in each foot, 6 areas for erosions and 6 areas for joint space narrowing. An overal l SvH is calculated by the summation of joint narrowing and joint erosion in the hands and the feet as measured by a radiologist or any other expert reviewer of radiographic images. Goals and dataset The RA2 DREAM chal lenge aims at faci l itating the creation of an automated method that quickly and accurately quanti f ies the degree of RA-associated joint damage in hands and feet of patients. Participants to the chal lenge wi l l use existing large sets of radiographs and scores from previous NIH-funded CLEAR and TETRAD registries. One of the technical chal lenges that participants need to address is the presence of wedding rings and other jewelry on the hands of patients included in the dataset. Based on the number of f i lms including jewelry, it looks to be common practice to al low the patient to RA2 DREAM Challenge