Kenny Walter is an editor with HCPLive. Prior to joining MJH Life Sciences in 2019, he worked as a digital reporter covering nanotechnology, life sciences, material science and more with R&D Magazine. He graduated with a degree in journalism from Temple University in 2008 and began his career as a local reporter for a chain of weekly newspapers based on the Jersey shore. When not working, he enjoys going to the beach and enjoying the shore in the summer and watching North Carolina Tar Heel basketball in the winter.
In data presented at AAO 2020, researchers discover new methods for detecting different underlying conditions.
Using fundus imaging, new research suggests clinicians could begin to detect a number of underlying conditions.
A team, led by Young Dae Kim, MD, Abbott Northwestern Hospital, developed and assessed convolutional neural network (CNN) models that can detect underlying hypertension, diabetes, and smoking status from retinal fundus images in data presented at American Academy of Ophthalmology (AAO) 2020 virtual conference.
From the normal retinal fundus images found in the SBRIA database, the investigators collected images of patients with hypertension, diabetes, and smoking habits. The research team matched 1 normal image from patients without hypertension, diabetes, and smoking habit to each identified image.
Next, the researchers developed disease detection models using ResNet-152.
Overall, 202,350 fundus images were included in the analysis, while the areas under the curve (AUCs) were 0.9782 in hypertension, 0.9894 in diabetes, and 0.7509 in smokers.
“Our CNN models provide excellent disease detection in test-sets with hypertension and diabetes, but not in smokers,” the authors wrote. “Our study suggested that there is a change in hypertension and diabetes even in the normal fundus images, and showed that fundus images may be more helpful in health care. Cross-validation and further studies will be needed.”
Last year, researchers found an individual’s diet could play a large role in the risk of developing age-related macular degeneration than previously thought.
Results of a study from the State University of New York at Buffalo revealed patients consuming a diet high in red and processed meats, refined grains, fried food, and high-fat dairy were 3 times more likely to develop the retinal condition.
“What we observed in this study was that people who had no AMD or early AMD at the start of our study and reported frequently consuming unhealthy foods were more likely to develop vision-threatening, late-stage disease approximately 18 years later,” said investigator Amy Millen, PhD, associate professor and associate chair of epidemiology and environmental health at University at Buffalo, in a press release.
One of the leading causes of vision loss among Americans, investigators sought to determine how diet plays a role in the development of late-stage age-related macular degeneration. To evaluate how a “Western” dietary pattern, which investigators classified as high in consumption of red and processed meat, fried food, refined grains and high-fat dairy, impacted disease development, investigators conducted an analysis using data from the Atherosclerosis Risk in Communities (ARIC) Study.
Upon analyses, there were no observed associations between either dietary pattern examined and incident any or incident early age-related macular degeneration.
Conversely, when examining the development of late age-related macular degeneration, investigators found participants with a Western pattern score were at 3 times greater risk of developing the condition compared to those below the median (OR = 3.44; 95% CI, 1.33-8.87; P-trend = 0.014). Investigators also noted the risk of developing late age-related macular degeneration was decreased but not statistically significant among patients with a prudent pattern score above compared to those with a score below the median (OR, 0.51; 95% CI, 0.22-1.18; P-trend =0.054).
The study, “Detection of Underlying Hypertension, Diabetes and Smoking Status From Retinal Fundus Images Using Deep Neural Network,” was published online by AAO 2020.