A New Infrared Thermography Based Approach to Ankylosing Spondylitis Assessment: A Preliminary Study

Area of Science:

• Biomedical Engineering
• Medical Imaging
• Rheumatology

Background:

• Traditional ankylosing spondylitis (AS) assessment requires costly hospital visits.
• There is a need for convenient, non-contact methods for monitoring AS.

Purpose of the Study:

• To develop and validate a non-contact, home-based method for AS assessment using infrared thermal imaging.
• To classify AS status (healthy, inactive, active) and correlate thermal data with inflammatory markers.

Main Methods:

• An Adaptive Weight U-Net (AWU) framework was developed to analyze knee infrared thermal images.
• The framework computed temperature difference ($\Delta T$) between knee joint and peripheral skin.
• Kendall's correlation analysis assessed the relationship between $\Delta T$, C-reactive protein (CRP), and erythrocyte sedimentation rate (ESR).

Main Results:

• The AWU framework achieved 77.38% classification accuracy for AS status in 221 participants.
• Perfect discrimination was observed between healthy individuals and those with active AS.
• A moderate positive correlation was found between $\Delta T$ and CRP ($\tau =0.5920$), and a weak positive correlation with ESR ($\tau =0.4236$).

Conclusions:

• Infrared thermography, analyzed via the AWU framework, shows promise as a complementary tool for AS assessment.
• This non-contact, home-based method could facilitate remote monitoring and reduce patient burden.
• The findings support the use of thermal imaging for evaluating knee inflammation in AS patients.