Facebook says AI can predict four days in advance if a coronavirus patient’s condition will deteriorate – just a single chest x-ray
- Facebook and NYU have built AI to predict developments among patients with COVID
- IT has been trained with thousands of chest x-rays with and without COVID
- The software is based on a series of three machine learning models
- One predicts damage with a scan, and another uses a sequence of scans
- The third model is able to predict whether additional oxygen will be needed
Facebook claims to have designed software that can predict whether a coronavirus patient’s health will deteriorate or need oxygen just by scanning a chest x-ray.
Collaborating with New York University (NYU), the social media company says the system can calculate these developments for four days.
Together they have built three machine learning models to help doctors better prepare as cases around the world continue to grow.
One model is designed to predict damage using a single chest x-ray, another does the same thing, but through a series of X-rays, and the third uses an x-ray to determine if and how much extra oxygen a patient may need.
Scroll down for the video
Facebook and NYU have built three machine learning models to help doctors better prepare as cases around the world continue to grow. One model is designed to predict damage using a single chest x-ray, another does the same thing, but through a series of X-rays
Coronavirus has been a strong force on the world for more than a year, starting in Wuhan, China.
There are over 93 million cases worldwide and about two million people have lost their lives to the virus.
However, the United States weighed in at over 23 million cases and over 389,000 deaths.
As cases continue to grow, scientists and technology companies are working on innovative ways to capture symptoms before they take a deadly turn.
The third uses an x-ray to determine if and how much extra oxygen a patient may need. AI can make predictions four days in advance based on X-rays.
Facebook and NYU each have five researchers who have worked on coronavirus prediction software.
“Our model using sequential chest X-rays can predict up to four days (96 hours) in advance whether a patient may need intensive care solutions, generally exceeding the predictions of human experts,” the authors said in -a blog post published on Friday.
The team first “pre-trained” the model with a public chest X-ray database on MIMIC-CXR-JPG and CheXpert.
The system was then instructed on NYU coronavirus chest radiographs, which included a total of 26,383 images from 4,914 patients, which allowed the models to label a patient’s condition worsening by 24, 48, 72 or 96 hours concerned.
“Based on studies by readers with radiologists at NYU Langone, our models using X-ray imaging sequences exceeded the performance of human experts in predicting ICU needs and predicting mortality and general long-term adverse event predictions (up to 96 hours), ‘the team shares on the blog.
Being able to predict whether a patient will need oxygen resources would also be a first and could help hospitals as they decide how to allocate resources in the coming weeks and months.
Facebook claims to have designed software that can predict whether a coronavirus patient’s health will deteriorate or need oxygen just by scanning a chest x-ray
William Moore, MD, professor of radiology at NYU Langone Health, shares in a statement: “We have been able to show that by using this AI algorithm, serial chest radiographs can predict the need to escalate care in patients with COVID-19. ‘
As COVID-19 continues to be a major public health issue, the ability to predict a patient’s need for increased care – for example, admission to the ICU – will be essential for hospitals.