- Using artificial intelligence (AI) and machine learning a new
computational tool was developed that can screen patients for eye
- AI-based system was trained using non-invasive
eye scans conducted with optical coherence tomography.
- The machine could tell if a patient should be referred for
treatment within 30 seconds, with more than 95% accuracy.
A new computational tool that can
screen eye diseases efficiently was developed using artificial intelligence and
machine learning techniques. The tool developed by research teams at Shiley Eye
Institute at UC San Diego Health and University of California San Diego School
of Medicine, with colleagues in China, Germany and Texas, could potentially speed up the diagnosis
and eventually treatment of common but blinding eye diseases. The study is
published in the journal Cell
Artificial intelligence (AI)
is a technology of
the future with immense potential in all fields including medicine. Today, in
its early stages the approach is laborious, expensive, and requires using
millions of data to train an AI system. However, the new study has developed a
feasible and efficient way of training the AI system.
‘The artificial intelligence-based computational tool diagnoses blinding retinal diseases within 30 seconds with 95% accuracy.’
intelligence has huge potential to revolutionize disease diagnosis and
management by doing analyses and classifications involving immense amounts of
data that are difficult for human experts - and doing them rapidly," said
senior author Kang Zhang, MD, PhD, professor of ophthalmology at Shiley Eye
Institute and founding director of the Institute for Genomic Medicine at UC San
Diego School of Medicine.
The AI-based neural network was developed using a
technique called transfer learning. Here the knowledge gained in solving one
problem is stored by a computer and applied to different but related problems.
For example, if an
AI neural network optimized to recognize anatomical parts of the eye like the
retina or the cornea, the system can more quickly and efficiently identify and
evaluate these structures when examining images of a whole eye. This allows the
system to learn effectively with a much smaller dataset.
Efficiency of the AI-based System
This study was
primarily focused on using the
new tool to diagnose two common causes of irreversible blindness: macular
degeneration and diabetic macular edema, both if detected early are treatable.
The AI-based tool's diagnoses of eye scans were
compared with diagnoses from five ophthalmologists who reviewed the same scans.
The machine was found to perform alike any well-trained ophthalmologist. The
machine could decide if a patient should be referred for treatment within 30
seconds and that too with a 95% accuracy.
from eye diseases, the AI tool was also used to diagnose childhood
Based on the chest
X-rays, the system could differentiate between viral and bacterial pneumonia
with greater than 90 percent accuracy.
The new approach
has great potential in early detection of diseases, which ultimately leads to early treatment. The
AI-based system is also
a great resource for regions with a scarcity of medical professionals and
- Artificial intelligence quickly and accurately diagnoses eye diseases and pneumonia - (https://www.eurekalert.org/emb_releases/2018-02/uoc--aiq021518.php)
- Cell, Kermany et al. "Identifying Medical Diagnoses and Treatable Diseases by Image-Based Deep Learning." Cell, (2018) DOI: 10.1016/j.cell.2018.02.010