Clicky

AI-based technology to identify patients at risk of serious illness as a result of blood infections

A new technology developed at Tel Aviv University will make it possible to use artificial intelligence to identify patients who are at risk of serious illness from blood infections.

The researchers trained the AI program to examine the electronic medical records of approximately 8,000 patients at Ichilov Hospital in Tel Aviv who were found positive for blood infections. These records included demographics, blood test results, medical history, and diagnosis.

After studying the data and medical history of each patient, the program was able to automatically identify the risk factors in the medical records with an accuracy of 82%. According to the researchers, this model could even serve as an early warning system for doctors in the future, allowing them to classify patients according to their risk for serious illnesses.

Behind this groundbreaking research with the potential to save many lives are students Yazeed Zoabi and Dan Lahav from the laboratory of Prof. Noam Shomron of the Sackler Medical Faculty of Tel Aviv University in collaboration with Dr. Ahuva Weiss Meilik, Head of the I-Medata KI Center at the Ichilov Hospital, Prof. Amos Adler and Dr. Orli Kehat. The results of the study were published in the journal Scientific reports.

The researchers explain that blood infections are a leading cause of morbidity and mortality worldwide. Therefore, it is very important to identify the risk factors for developing serious illness in the early stages of infection with a bacterium or fungus. Most of the time the blood system is sterile, but infection with a bacterium or fungus can occur during surgery or as a result of complications from other infections such as pneumonia or meningitis. The diagnosis of infection is made by taking a blood culture and transferring it to a nutrient medium for bacteria and fungi. The body’s immunological response to infection can cause sepsis or shock, dangerous conditions with a high mortality rate.

We worked with the medical records of approximately 8,000 patients at Ichilov Hospital who were found positive for blood infections between 2014 and 2020, during their hospital stay and up to 30 days after, whether the patient died or not. We entered the medical records into software based on artificial intelligence; We wanted to see if the AI could see information patterns in the files that would allow us to automatically predict which patients would develop serious illness or even death as a result of the infection. “

Noam Shomron, Professor, Sackler Faculty of Medicine, Tel Aviv University

To the satisfaction of the researchers, after their training, AI achieved an accuracy of 82% in predicting disease progression, even ignoring obvious factors such as patient age and the number of hospital stays suffered. After the researchers entered the patient data, the algorithm was able to predict the course of the disease, which suggests that it will be possible in the future to classify patients according to their health risk – in advance.

“With artificial intelligence, the algorithm was able to find patterns that surprised us, parameters in the blood that we hadn’t even thought of,” says Prof. Shomron. “We are now working with healthcare professionals to understand how this information can be used to grade patients according to the severity of the infection.

Since the success of the study, Ramot, Tel Aviv University’s technology transfer company, has been working to apply for a worldwide patent for the groundbreaking technology. Keren Primor Cohen, CEO of Ramot, said: “Ramot believes in the ability of this innovative technology to transform the early detection of patients at risk and help hospitals reduce costs. This is an example of the effective collaboration between university researchers and hospitals, improving the quality of medical care in Israel and around the world. “

Source:

Journal reference:

Zoabi, Y., et al. (2021) Predicting the Outcome of a Bloodstream Infection Using Machine Learning. Scientific reports. doi.org/10.1038/s41598-021-99105-2.