A deep learning artificial intelligence (AI)-method accurately predicted the development of rheumatoid arthritis (RA) at an early stage using MRI scans.
Predicting early RA from MRI images can help initiate prompt treatment, possibly preventing the chronicity of the disease. Visual scoring (eg, with the RAMRIS scoring system) from extremity MRI scans has been used to manually identify key risk factors in RA. At EULAR 2023, held May 31 to June 3, in Italy, Yanli Li presented the findings of a study that sought to whether AI interpretations of MRI images could provide more accurate predictions than visual scoring by medical professionals.
The model was first trained to understand anatomy from MRIs of wrists and metacarpophalangeal joints of healthy controls. In a second step, it learned to distinguish between the different groups (patients with clinically suspect arthralgia [CSA] versus healthy controls and early-onset arthritis versus healthy controls). In a third step, the AI was taught to distinguish RA from other arthritides. Finally, the system had to predict RA development in 2 years in patients with CSA. The model’s accuracy was evaluated with the area under the receiver operator curve (AUC).
The AI analyzed MRI scans from 1,974 people with either early-onset arthritis or CSA, of whom 651 went on to develop RA. On the test set, the proposed model obtained a mean AUC of 0.683 in the early-onset arthritis group and 0.727 in the CSA group. The AI performed about as accurately as experts using RAMRIS.
As Li emphasized during the presentation, AI-based RA prediction is reliable as it looks at known inflammatory signs and features, such as synovial inflammation. Moreover, the self-learning AI system showed similar efficacy for scans of either wrists or feet. The system can be further improved with more clinical data.
Copyright ©2023 Medicom Medical Publishers
