Publications by authors named "Michiko Yonehara"

Article Synopsis
  • Artificial intelligence shows potential for diagnosing allergic conjunctival diseases (ACDs), but its development is hindered by a lack of tailored image databases and model explainability.
  • This study utilized a dataset of 4,942 slit-lamp images from 10 institutions in Japan to create an AI model that diagnoses ACDs and explains its reasoning.
  • Results indicated the AI model achieved a diagnostic accuracy of 86.2%, significantly outperforming board-certified ophthalmologists, and demonstrated effective identification of key clinical indicators for various ACDs.
View Article and Find Full Text PDF