bodyType

customContents

entityId

entityUuid

articlePageQuery

filter

article

articleBaseAdvertisingInfoArticle

articleListArticle

articleSeoArticle

helpersCheckIsPostiesArticle

hooksArticleAcknowledgementGateOverlayArticle

hooksTrackPageViewArticle

hooksContentInterestArticle

articleLinkingDataContent

hooksAmpRedirectArticle

articleSchemaContent

liveArticle

advertZone

sentiment

contentLock

topics

types

sections

moreOnThisArticles

urlAlias

commentCount

authorLocations

authors

corrections

sponsorType

displaySlideShow

multimediaEmbed

printHeadline

seriesContainer

socialHeadline

headline

flag

firstTopic

glossary

flattenTypeIds

publishedDate

additionalInfoBoxes

quizIds

images

image

relatedLinks

relatedNewsletters

__typename

createdDate

sponsor

paywallTypes

pdfFiles

seriesReverseOrder

series

disableOffPlatformContent

published

updatedDate

summary

subHeadline

body

keywords

readingTime

customTimezone

liveContents

LO WEI

A computer program that scans images of patients' eyes may offer a cheap and speedy way to identify people at risk of a stroke.
Developed by Chinese University researchers, the system automatically checks blood vessels from retinal images to identify potential stroke victims.
Tests have shown it to be up to 90 per cent accurate and it may go into widespread use after three more years of study.
"Our aim is to popularise the system so that more people can be detected before they get a stroke," said Professor Benny Zee Chung-ying, head of the university's biostatistics division.
Images from retinal cameras are uploaded to the system, which gives an analysis of the patient's blood vessel patterns and whether there is any narrowing or obstruction.
Doctors can then warn patients of the possible risk and recommend medication or lifestyle changes, such as quitting smoking.
The main method available now for stroke risk detection is the MRI scan, but low capacity and high cost limit its use.
Past researchers' efforts to use retinal images for the purpose are recorded in medical literature.
But Zee said the biggest difficulty was that it takes a long time to make out blood vessel properties by looking at the images.
He said the new system was quicker and required less manpower. It can also diagnose diabetic retinopathy, a complication among diabetes patients.
Zee said the team was working on applying the system to detect the risk of other diseases including Alzheimer's, diabetes and age-related macular degeneration, a vision problem common in the elderly.
The researchers have completed a study on 244 patients aged 50 to 70, of whom half have had a stroke and half have not.
It found that the system identifies those who have had a stroke with 90 per cent accuracy.
It missed the other 10 per cent mainly because they did not show common retinal blood vessel characteristics.
The researchers are now recruiting 1,200 more patients over 65 who have no stroke history to be tested with the system.
The results will be compared with their MRI scan results to further check the system's accuracy.
The World Stroke Organisation estimates one in six people worldwide will have a stroke.
In Hong Kong, about 18,000 people suffer first or recurrent strokes each year and 20 per cent of them die.


Eye scanner that can detect stroke risk developed by Chinese University

The retinal image analysis system. Photo: Thomas Yau

News

Hong Kong

Researchers at Chinese University develop program that can analyse retinal images
