Photo: VisionQuest Biomedical
New technology could prevent widespread misdiagnosis of cerebral malaria in low-income countries, helping more children receive accurate diagnoses of life-threatening conditions, according to an American company that is developing a tool for this.
VisionQuest Biomedical Inc. told Devex that its tool, which is capable of diagnosing cerebral malaria via the retina, could be used in remote areas and places where there is a shortage of eye care professionals, and has plans to roll it out in African countries.
In regions across Africa where malaria is widespread, children who are rushed to health facilities in a coma are often misdiagnosed with cerebral malaria, a severe form of malaria that can cause brain damage, as well as seizures and coma. On the African continent, cerebral malaria primarily impacts children.
But one of the problems is that the symptoms of cerebral malaria overlap with other diseases and children can have the malaria parasite present in their blood with no symptoms. That means a blood test could give a positive result for malaria, even when that is not the primary cause of a child’s illness. In reality, they could be in a coma for other reasons, such as meningitis, which left untreated can be fatal or cause lasting neurological damage, experts said.
“Making an accurate diagnosis of cerebral malaria is a critical challenge,” said Sidney Ogolla, senior research officer at the Kenya Medical Research Institute.
An autopsy study in Malawi found that 25% of diagnoses of cerebral malaria were wrong, and the patient’s cause of death was something else. But there are clues in a patient’s eyes: Retinal abnormalities can help determine with high levels of accuracy if a patient has cerebral malaria.
Though such clues could significantly lower the number of misdiagnosed cases, challenges still exist in low-resource settings where having an ophthalmologist on-hand to do the screenings is rare. For example, in Tanzania, there is less than 1 ophthalmologist for every 1 million residents.
In response to this challenge, VisionQuest, which focuses on using artificial intelligence tools to detect diseases using the retina, is developing a portable and automated tool so these screenings can be carried out by someone with minimal training and in remote areas.
“These tools will allow us to make a better diagnosis to catch those children that may test positive for malaria, but that's not the main cause why they're in the hospital,” said Simon Barriga, chief executive officer at the company. “We potentially can save a lot of children from dying from an incorrect diagnosis.”
Researchers are currently involved in clinical research studies using the software in three African countries to help strengthen the algorithm. The company said it plans to start registering it in Malawi, Kenya, Zambia, Rwanda, and Nigeria next year. It estimates that the product will be commercially available in those countries in 2022, with plans to roll it out in other countries following this.
The research that identified these clues in the eyes began in the early 1990s, in Blantyre, Malawi, when two friends met for dinner. Dr. Susan Lewallen, an ophthalmologist, and Dr. Terrie Taylor, a doctor who coordinates a malaria research program, began to discuss the mysteries around the fast progression of cerebral malaria in the children Taylor was treating and the unknowns surrounding how the disease killed them.
Lewallen asked Taylor if they had ever dilated the eyes of these children. The retina is, in some ways, an extension of the brain, and can give clues into what’s happening in the head. Taylor’s team hadn’t done this yet in detail.
Out of curiosity, Lewallen started to examine the retina of the children at Taylor’s center. What she found surprised her. She saw blood vessels that were white, instead of red, and cloudy retinal patches.
It became clear that no researcher had yet studied this issue systematically on children. The two embarked on an autopsy study in 1996, funded by the U.S. National Institutes of Health. They examined the eyes of children who died after a cerebral malaria diagnosis, comparing them with the brains.
“Everything became clear in the context of the autopsy study. It turned out that not all of the kids who we thought had cerebral malaria did,” Taylor said.
In 2015, Lewallen and Taylor were asked by Dr. Vinayak Joshi, principal investigator of the cerebral malaria project at VisionQuest Biomedical, to help the company develop an algorithm to detect cerebral malaria through a retinal scan. These efforts, in coordination with the University of Liverpool, are funded through a grant from the U.S. National Institutes of Health.
They have since developed software that can be used with an optical camera system and a smartphone. It only requires a few days training to use, according to Joshi.
The price for the camera, which the health facility would need to buy seperately, ranges from $1,500 to $2,000. Then, health care providers would pay about $500 each year for the VisonQuest Biomedical software, Joshi said. The camera can also be used to detect other eye diseases, such as diabetic retinopathy, using other software tools. The company hopes that philanthropic organizations and NGOs will assist in financing the software’s distribution in low-resource environments.
Chulaimbo County hospital, outside of Kisumu, in western Kenya, is one of the locations where researchers are currently testing out the product and refining the algorithm — scanning the retinas of children diagnosed with cerebral malaria, after other illnesses have been ruled out.
In a normal clinical setting, conducting a number of tests to rule out other diseases is challenging and can take several days, Kenya Medical Research Institute’s Ogolla said.
“But with this system, it's instant,” he said. “We expect ultimately that this system will be a simplified approach of improving the diagnostic accuracy and will provide a bedside tool for confirming cases of cerebral malaria.”
The Collateral Damage of “Non-Essential” Procedure Designations
Matt Birnholz, MDPeer