Algorithm Spots COVID-19 Cases from Eye Images: PreprintSeptember 27, 2020
cientists describe a possible screening method for COVID-19 supported eye images analyzed by AI . Scanning a group of images from several hundred individuals with and without COVID-19, the tool accurately diagnosed coronavirus infections quite 90 percent of the time, the developers reported during a preprint posted to medRxiv September 10.
“Our model is sort of fast,” Yanwei Fu, a scientist at Fudan University in Shanghai, China, who led the study, tells The Scientist. “In but a second it can check results.”
Currently, screening for coronavirus infection involves CT imaging of the lungs or analyzing samples from the nose or throat, both of which take time and need professional effort. A system supported a couple of images of the eyes that would triage or maybe diagnose people would save on both costs and time, says Fu. But the investigation by Fu’s team is preliminary and both ophthalmologists and AI specialists say they’d want to ascertain far more information on the technique—and its performance—before being convinced it could work.
Volunteers at Shanghai Public Health Clinical Centre in Fudan each had five photos of their eyes taken using common CCD or CMOS cameras. Of 303 patients, 104 had COVID-19, 131 had other pulmonary conditions, and 68 had eye diseases. A neural network tool extracted and quantified the features from different regions of the attention and an algorithm recognized the ocular characteristics of every disease. A neural network may be a series of algorithms for solving AI problems, learning because it goes along during a way that mimics the human brain. The researchers then administered a validation experiment on alittle dataset from healthy people, COVID-19 patients, pulmonary patients, and ocular patients.
Of 24 people with confirmed coronavirus infections, the tool correctly diagnosed 23, Fu tells The Scientist. and therefore the algorithm accurately identified 30 out of 30 uninfected individuals.
Coronavirus infections, not just those caused by SARS-CoV-2, have long had associations with the attention , causing inflammation of the transparent membrane that covers the within of the eyelid and whites of the eyeball, a condition called conjunctivitis, or pink eye. The eyes also offer a route to infection for respiratory viruses, including coronaviruses.
Human coronavirus NL63, which causes cold symptoms, was first identified in 2004 during a baby with bronchiolitis and conjunctivitis. Subsequent studies showed that a minority of youngsters infected with this coronavirus suffer from this condition .
Although conjunctivitis remains a possible symptom of coronavirus infections, but 5 percent of COVID-19 patients actually present with eye symptoms, notes Daniel Ting, ophthalmologist at the Singapore National Eye Centre, who has published on this subject and deep learning in ophthalmology. “If you look to develop an AI system to detect COVID-19 supported [limited numbers of] eye images, i feel the performance isn’t getting to be great,” especially given the low prevalence of eye symptoms. He doubts the performance of the algorithm also because “a lot of eye manifestations might be thanks to reasons aside from COVID-19.”
Ting cautions that the sample size of 303 patients and 136 healthy individuals within the Shanghai study is just too small to draw strong conclusions. “To develop an honest deep learning system to automatically detect some unique features from any medical imaging requires more patients,” he says. “In order to extend the reliability of this study, an equivalent size would wish to be multiplied by a minimum of ten times, so, thousands of patients.”
Fu has started down this road, increasing the amount of participants and broadening the kinds of subjects. “We are now doing more double-blind tests within the hospitals, with patients, some with eye diseases,” he says. The group also plans to introduce a web screening platform that uses the algorithm to screen for COVID-19.
“As an ophthalmologist it might be very surprising if there’s a definite COVID viral conjunctivitis pattern as against other similar sorts of viral conjunctivitis,” ophthalmologist Alastair Denniston, the director of the Health Data Research Hub for Eye Health in Birmingham, UK, writes in an email to The Scientist. “This is unlike building an algorithm for conditions which are biologically more distinct like degeneration ,” he writes.
He notes that if there have been a singular pattern evident in COVID-19 cases, “then the comparison for training and testing should be against cases that look similar,” like non–COVID-19 viral conjunctivitis or other causes of a red eye related to colds caused by adenovirus or rhinovirus. He also faults the paper in not providing “the necessary description to actually critique the science in terms of how they built and (tried to) validate the model.”