Robots in Rheumatoid Arthritis treatment

Rheumatoid arthritis (RA) is a chronic inflammatory disease affecting around 1% of the world’s population. The disease causes swelling and pain in the joints, and if left untreated, it can cause stiffness, joint destruction, and disabilities. RA can lead to reduced quality of life and diminished work capacity.
RA has become more widespread among the elderly, and with the worldwide population increasingly aging, RA is soon going to present a larger problem for a health care system with an already severe shortage of rheumatologists and no prospects of increasing hospital budgets.
Early detection, followed by monitoring, plays a key role in managing and treating the disease. Ultrasound examination of joints has proven to be a sensitive imaging tool for detecting signs of early disease and evaluate arthritis activity in established disease. Unfortunately, ultrasound examinations have a long waiting list that delays the patients’ treatments.
RoPCA’s ultrasound platform ARTHUR and the artificial intelligence DIANA present a solution, both to the increasing number of RA patients, the hospital cost, and the lack of specialists.


ARTHUR (Arthritis Ultrasound Robot) is the automated ultrasound platform for scanning rheumatoid arthritis patients. The platform automatically captures ultrasound images of fingers and wrist joints and evaluates the image quality using deep learning. The images are then evaluated for disease activity by DIANA, which sends the images with signs of disease onward for a rheumatologist to assess.
Conventional ultrasound examinations are dependent on manual labor and ultrasound diagnostics are time-consuming for doctors and they therefore present a high cost for hospitals. Patients can as a result look forward to long waiting lists. Since ARTHUR isn’t operator dependent, as it interacts directly with the patient without the need of a doctor, it can allocate more time doctors can use elsewhere.
The long waiting list for patients under investigation for RA or with an established case of RA, means there is a risk of missing early disease, and for the patients with confirmed RA, too much time passes between the disease control checkups. When ARTHUR takes care of scanning the patients, the doctors can better prioritize which patients to see first, and thereby reduce the waiting time.
RA drugs, which are extremely expensive, are increasingly being used and frequent monitoring is consequently needed to adjust the treatments. ARTHUR’s patient monitoring ensures that ineffective treatment gets replaced fast or that treatment gets reduced if the patient is in remission.
Ultrasound examinations suffers from a central problem. The quality of the ultrasound scanning is operator dependent and difficult to repeat, which causes variability in the scanning results. This irregularity can influence the monitoring of the disease progression. The ARTHUR platform is capable of automatic ultrasound scanning that ensures a systematic and uniform method to perform ultrasound scans every time.


DIANA (Diagnosis Aid Network for Rheumatoid Arthritis) is the assistant tool for evaluating and monitoring the disease activity in the joints scanned by ARTHUR. The DIANA software system can be used in collaboration with ARTHUR, where the combined system will allow for automatic scanning, analysis and report generation. The reports can be used by doctors in the diagnosing and monitoring of patients with RA. DIANA can also be used as a standalone system used on manually captured ultrasound images.
The DIANA software system is based on Convolutional Neural Networks (CNN’s), which are state-of-the-art algorithms used for image analysis. The networks have been trained on thousands of ultrasound images, which have been acquired with the assistance of expert rheumatologists, making the system able to classify the degree of RA. The ultrasound images used for training, has been labeled by expert rheumatologists according to the EULAR-OMERACT standards for diagnosis of rheumatoid arthritis.
RoPCA’s dedicated AI team is continuously working on improving DIANA to optimize performance and to increase transparency in the prediction pipeline. RoPCA works closely with experts in this field to achieve this vision
AI segmentering
* Areas marked by AI

Do you want to know more?

Documents and Manuals

Please enter the provided password.

If you do not have a password and wish to access this content, please use the contact form.