Artificial Intelligence in Thyroid Eye Disease Imaging: A Systematic Review

Strategic Insights -

Researchers from Shanghai Jiao Tong University School of Medicine have conducted a systematic review of artificial intelligence (AI) applications in thyroid eye disease (TED) imaging. The study aimed to characterize the research landscape, key challenges, and future directions of AI in TED diagnosis and treatment. The researchers analyzed 41 studies covering various AI applications, including diagnosis, activity assessment, severity grading, and treatment prediction, with a focus on CT, facial, and retinal imaging.

Key Takeaways:

  • The study found that AI-based imaging has strong potential to improve diagnostic accuracy and guide personalized treatment strategies in TED.
  • The majority of the studies were of moderate quality, and most employed deep-learning models such as residual network (ResNet) and Visual Geometry Group (VGG).
  • CT and facial imaging were the most common modalities, reported in 16 and 13 articles, respectively.
  • Researchers primarily used ResNet and VGG models for image classification and segmentation.
  • The study suggested that future research should prioritize robust study designs, the creation of public datasets, multimodal imaging integration, and interdisciplinary collaboration to accelerate clinical translation.

Statistics:

  • The study analyzed 41 studies covering various AI applications in TED imaging.
  • The sample sizes ranged from 33 to 2,288 participants, predominantly East Asian.
  • The majority of the studies (63%) were conducted in East Asia.
  • The most common AI models used were residual network (ResNet) and Visual Geometry Group (VGG).
  • The study found that image-based AI shows strong potential to improve diagnostic accuracy in TED.

Sources:

  • "Artificial intelligence in thyroid eye disease imaging: A systematic review" published in Survey of Ophthalmology, 2025.
  • Shanghai Jiao Tong University School of Medicine Reports Findings in Personalized Medicine (Artificial intelligence in thyroid eye disease imaging: A systematic review) published in Journal of Engineering, August 4, 2025.