Artificial intelligence at the service of kidney damage



[ad_1]

Thanks to the work of American researchers, an artificial intelligence has been developed. The latter can identify kidney damage in intensive care patients, 24 hours ahead.

An artificial intelligence that can detect kidney damage

Scientists from Northwestern University and the University of Texas Health Science Center have created a new artificial intelligence (AI). It can predict acute kidney damage 24 hours early in ICU patients. Indeed, this type of patient is more likely to develop this pathology that can destroy the renal system and become fatal if it is not detected early enough.

Scientists worked with a powered AI with clinical notes from 40,000 patient records. They integrated different data such as bad, ethnicity, age of patients, and also serum creatine levels recorded during the 72 hours preceding the examination. Thanks to all this information, artificial intelligence has been able to obtain good results.

For example, during the tests, she identified 50% of the cases at risk of having a kidney lesion within 24 hours of her examination. This new method is therefore considered quite reliable compared to other methods already known. And this AI will be able to further improve its results in the future, thanks to the development of its database by integrating the clinical notes of other patients. She will be able to refine her medical predictions to achieve a much higher percentage of success.

Share Article

[ad_2]
Source link