Various automation procedures relying on artificial intelligence and machine learning have been tested to diagnose age-related macular degeneration (AMD). This article, published in the International Journal of Imaging Systems and Technology, examined the automated diagnoses of affected eyes using an algorithmic approach.
The imaging data used for testing the automated diagnoses came from the STARE and AFIO datasets. The purpose of the approach used was to provide increased accuracy for early detection of AMD. The algorithm distinguishes the macular region and then determines if the macula is normal or abnormal based on the textural pattern, edge, and structural properties of the macular image.
Ultimately, support vector machine-based categorization was found to achieve high diagnostic accuracy. The overall accuracy of the automated system was 95% for the STARE database and 92% for the AFIO database [1].
Source:
[1] Khalid, S., Akram, M. U., Shehryar, T., Ahmed, W., Sadiq, M., Manzoor, M., & Nosheen, N. (2020). Automated diagnosis system for age‐related macular degeneration using hybrid features set from fundus images. International Journal of Imaging Systems and Technology, 31(1), 236–252. https://doi.org/10.1002/ima.22456
