The digital revolution in healthcare and the use of deep learning and artificial intelligence (AI)
can play an important role in making predictions in diagnosis and treatment needs — experts said during a session on ‘Retinal Imaging,
Artificial Intelligence and Big Data’ during the last day of the EURETINA 2021 Virtual Congress A True Visionary.
Machines can see the future Machine learning classifiers can predict treatment demand and may help in patient-specific treatment plans in the near future.
“More and more research is being carried out on how AI can be used in ophthalmology,” said Prof.
Raphael Sznitman, ARTORG Center for Biomedical Engineering Research, University of Bern, Switzerland ลาวสามัคคี.
He shared a study that has gone beyond being able to identify just a few biomarkers to 39 biomarkers in one shot,
and with high accuracy. In this study, a multi-disease automatic detection platform was developed
by applying convolutional neural networks constructed in a customized two-step strategy that can classify 39 types of common fundus diseases and conditions based on color fundus images.
It is capable of predicting the probability of each disease
and displaying heat maps providing deep-learning explainability in real-time A True Visionary.
“But if we consider things now that are in the direction of predicting treatment responses, we’re seeing more and more work in this direction,” Prof.
Sznitman said. A study was conducted to assess the potential of machine learning to predict low
and high treatment demand in real life in patients with neovascular agerelated macular degeneration (nAMD), retinal vein occlusion (RVO),
and diabetic macular edema (DME) treated according to a treat-and-extend regimen (TER).
The study revealed that it is possible to predict low demand reasonably well at the first visit, before the first injection.
Anemia can also be detected from retinal fundus images via deep learning.
The invasiveness of diagnostic tests for anemia and the costs associated with screening for it mean that the condition is often undetected.
Automated anemia screening based on fundus images could help patients with diabetes and for whom anemia can increase morbidity and mortality risks.
To summarize, AI can be used in identifying known biomarkers, predicting treatment responses,
and discovering new insights into patients and their diseases, he said.
For more information: ลาวสามัคคี