23-03-2020 3:26 am
Leave your thoughts
Physicians treating patients amid the coronavirus pandemic must make life or death decisions about who is being treated every day. That's more complicated than it sounds, and researchers in China are developing AI tools to help doctors make those choices.
But that raises another complicated question: should
artificial intelligence involved in medical decisions about life and death?
Doctors in the busy chaos of a pandemic hospital need to make clinical decisions quickly about treatment from one patient to another.
One argument is that a person with a better chance of fighting the disease is given resources rather than someone with a minimal chance of survival. The aim is to prevent the worst-case scenario that both people die: one because they had little chance, the other because their treatment was refused.
The other argument is that patients with the highest risk of death should be treated first.
Chinese researchers say they have come up with a piece of artificial intelligence technology that can help doctors make a more informed decision about who has the best chance of survival or the highest risk of death in competing Covid-19 patients.
Researchers from Huazhong University of Science and Technology (HUST) and Tongji Hospital in Wuhan, Hubei say they have developed an AI diagnostic tool that allows rapid analysis of blood samples to predict survival rates.
The developers claim that the AI tool achieved 90 percent accuracy in the mortality and survival rates of more than 400 patients based on blood samples collected from the date of admission to Tongji hospital.
They published the results of their ongoing research on the preprint server Medrxiv.org, a platform scientists around the world use to release non-peer-reviewed research on Covid-19.
The newspaper says the developers, led by Yuan Ye, a professor at the school of artificial intelligence and automation at HUST, are looking to improve system accuracy with a larger database in the near future.
Since patients are infected with the pneumonia caused by the coronavirus swamp hospitals around the world, the researchers say that with short time AI can help medical personnel with limited time and resources to decide which person will be treated first.
“There is currently no prognostic biomarker available to distinguish patients who require immediate medical attention and the associated death rate,” Yuan and colleagues wrote in their paper.
The goal of their AI system was to “identify high-risk patients before irreversible [lung] lesions occur,” she added.
When the Covid-19 outbreak was first discovered in Wuhan, central China, doctors and scientists knew little about the new virus that caused the disease.
Some patients with mild symptoms and no underlying health problems could suddenly fall into a critical condition. When they were rushed to an intensive care unit and given life support, fatal damage could have already been done.
The use of the new AI tool could allow “detection, early intervention, and potentially mortality reduction in patients at risk,” the researchers said.
AI technology is already being used in China to fight the pandemic. A
for example, has open public access to its AI diagnostic tool. This would allow doctors around the world to distinguish Covid-19 from other types of pneumonia in seconds by analyzing a patient's chest scan images.
Previous studies have also provided clues to the development of the disease in blood samples taken during routine body checks during a hospital stay.
But blood contains many chemicals, and it is all expensive and time consuming to check them all.
Yuan's team said they had identified three blood biomarkers that could carry the strongest signal of a Covid-19 infection. The lactate dehydrogenase (LDH) measured the level of lung damage, highly sensitive C-reactive protein (hs-CRP) indicated persistent inflammation and lymphocytes related to a decrease in immune cells.
A machine learning model based on these three biomarkers could precisely project the future development of the disease in a given patient, the researchers said.
According to Yuan, the accuracy of AI was influenced by when a patient's blood sample was taken, while later samples were more accurate. But he claimed that previous samples could still yield an accuracy rate of 90 percent or higher based on the patient's chance of survival.
The researchers said the tool had generated accurate forecasts for patients at Tongji hospital about 16 days in advance.
But a doctor who works in a Beijing public hospital and treats Covid-19 patients said the use of AI should be strictly monitored in early clinical assessment.
“This is a tool that can also be used to deny patients of old age or those with underlying conditions the right to be treated because a computer has decided they have almost no chance of survival,” said the doctor, requesting anonymity. .
He said it was also unclear whether the tool could be used outside of Wuhan.
An increasing number of studies suggest that the virus has mutated as it spread in China and worldwide, likely causing variations in the disease's development.
“Machine learning is a black box that is largely shaped by the data that is entered,” he said. “It may need to evolve just like the virus to adapt to different environments and people.”