Thermography + AI: A New Horizon in Rheumatoid Arthritis Management

Accurate detection of joint inflammation is the backbone of Rheumatoid Arthritis (RA) management. While tools like ultrasound or MRI are excellent, they can be time-consuming, operator-dependent, or not always readily available in daily clinical practice.

We are proud to share our most recent research published in RMD Open (BMJ), where we introduce an innovative, fast, and automated solution for this challenge.

📄 The Study: ThermoJIS

In this cross-sectional study, we recruited 595 participants, making it the largest thermography study in rheumatology to date. We developed and validated a Machine Learning-based computational method that analyzes thermal images of the hands to detect synovitis automatically.

🔍 Key Findings

• Correlation with the "Gold Standard": Our score, ThermoJIS, demonstrated a solid ability to detect active synovitis, validated against Power Doppler ultrasound (the current reference standard).
• Detecting the "Invisible": Most promisingly, the system was able to identify subclinical inflammation in patients who were theoretically in clinical remission but still showed underlying activity via ultrasound.
• Total Automation: Unlike subjective assessments, this algorithm automatically selects the regions of interest (ROI) and evaluates thermal patterns without human intervention, ensuring objectivity.

💡 Why does it matter?

This non-invasive and affordable method opens the door to objective assessments in seconds during a regular consultation. Furthermore, since it only requires a thermal camera (including those adaptable to smartphones), it points toward a promising future for remote monitoring and telemedicine, allowing patients to be monitored more frequently and precisely.

Congratulations to the entire team, including Isabel Morales-Ivorra, Javier Narváez, and Manuel Alejandro Marín-López, for this significant advancement at the intersection of technology and healthcare.

"This method provides a new horizon for clinicians to achieve objective, real-time data on joint inflammation, moving us closer to truly personalized rheumatology."

👇 Read the full article here: https://rmdopen.bmj.com/content/8/2/e002458

Reference: Morales-Ivorra I, Narváez J, Gómez-Vaquero C, et al. Assessment of inflammation in patients with rheumatoid arthritis using thermography and machine learning: a fast and automated technique. RMD Open 2022;8:e002458. doi: 10.1136/rmdopen-2022-002458