Close

International experts developed an artificial intelligence (AI) tool that can estimate how much extra oxygen a COVID-19 patient may require.

Medical Xpress said over 20 hospitals worldwide collaborated with NVIDIA, a pioneer in AI technology, to develop a novel AI-based approach known as federated learning using data from five continents.

Researchers published their study titled "Federated Learning for Predicting Clinical Outcomes in Patients With COVID-19" in the journal Nature Medicine.

How AI Predict COVID Patients' Oxygen Needs

News Medical said the method analyzes chest X-rays and electronic health data from hospitalized patients with COVID-19 symptoms using an algorithm.

Artificial Intelligence
(Photo: Possessed Photography on Unsplash)
Artificial Intelligence

Experts brought the analysis together to develop an AI tool - EXAM (electronic medical record (EMR) chest X-ray AI model) - that could forecast the oxygen demands of hospital Covid patients anywhere in the globe once the algorithm had "learned" from the data.

The findings revealed that it could estimate how much oxygen a patient will need within 24 hours of admission to the emergency room, with a sensitivity of 95 percent and a specificity of over 88 percent.

All patient data were anonymized to ensure absolute patient confidentiality, and experts gave an algorithm to each institution to ensure that no data was exchanged or left its original location.

By creating the EXAM model with federated learning and objective, multimodal data from multiple continents, researchers could create a generalizable model to assist frontline clinicians globally.

ALSO READ: Secret of Success Revealed: AI Study Unveils Magical Formula of Film Directors, Famous Artists Like Van Gogh


The research looked at the results of about 10,000 Covid patients from all around the world.

Professor Fiona Gilbert, the lead researcher at the University of Cambridge, believes that federated learning can bring AI innovation to the clinical workflow.

Dr. Mona G Flores, NVIDIA's Global Head for Medical AI, noted that federated learning allows researchers to cooperate and set a new benchmark for what AI can accomplish worldwide. According to Dr. Flores, the recent breakthrough will improve AI in healthcare and businesses trying to create powerful models without losing privacy.

How AI Could Further Help the Modern Healthcare

Business Insider said the world has seen what artificial intelligence can accomplish in the medical area. Scientists predict that artificial intelligence in healthcare will expand at a nearly 50% pace between 2017 and 2023.

Indeed, according to Fierce Biotech, the UK government is actively investigating AI's potential to prevent tens of thousands of cancer deaths per year by 2033.

BuiltIn also mentioned that a robot named Eve aided one of the most significant artificial intelligence achievements in medication research. According to the report, Eve found that a common toothpaste component might combat malaria-causing parasites.

Using IBM's AI, a team of researchers from the University of North Carolina could come up with exact therapies for over 1,000 patients. It may appear to be a modest number right now, but consider how many more lives may be saved as technology advances.

RELATED ARTICLE: Alzheimer's Disease Can Be Detected Early; New Technology Developed Shows 99-Percent Accuracy Rate

Check out more news and information on Medicine and Health in Science Times.