Artificial intelligence may assist in pinpointing AMD progression
BALTIMORE — A machine learning algorithm that predicts individual drusen regression using OCT-based biomarkers may predict when early or intermediate age-related macular degeneration will progress into severe AMD, according to a pilot study presented here at the Association for Research in Vision and Ophthalmology annual meeting.
“[The algorithm] would support ophthalmologists in estimating patients’ risk of developing severe AMD needed for patient counseling and would allow to adjust their screening schedule according to their risk,” Hrvoje Bogunovic, PhD, told Healio.com/OSN.
OCT images were obtained from 61 eyes of 38 patients with early/intermediate AMD every 3 months for at least 15 months. Researchers identified and characterized individual drusen and their development from baseline to the next follow-up. The point of regression of each druse was determined algorithmically.
Nine hundred thirty-eight drusen were identified, out of which 71 drusen (7.6%) regressed during follow-up.
“Detection of individual drusen regression within 12 months achieved an AUC = 0.76, corresponding to a sensitivity of 0.73 and a specificity of 0.70. The presence of [hyperreflective foci] was found to be an important biomarker for incoming drusen regression,” according to the presentation.
The algorithm will contribute to the development of new therapeutics that target slowing AMD progression, according to the researchers.
“Once validated it could serve as a clinical decision support, ie as another objective ‘pair of eyes’ that will look at the OCT images and their changes with time,” Bogunovic said.
The next step is to validate its performance on a larger cohort, he said. – by Abigail Sutton
Bogunovic H, et al. Personalized prognosis in early/intermediate age-related macular degeneration based on drusen regression. Presented at: Association for Research in Vision and Ophthalmology annual meeting; May 7-11, 2017; Baltimore.
Disclosure: Bogunovic reports no relevant financial disclosures. Please see the study for all other authors’ relevant financial disclosures.