A team of expert scientists who came together under the roof of Google has developed an artificial intelligence that can predict if patients will die within 24-48 hours using healthcare data.
Google’s team of computer scientists and medical scientists from the three largest universities in the United States claims that a patient in the hospital is the best way to predict whether a healthy individual will come out of the morgue as a healthy individual or as a deceased individual there.
This work is based on artificial intelligence, as it is in every research conducted by Google. Using a team of artificial intelligence software with advanced deep machine learning, the team explained that the technology they developed better analyzed patient data and predicted the results more successfully than other methods.
Artificial intelligence software uses patient demographics, previous diagnoses and treatments, laboratory results and vital findings to make the most accurate prediction about the patient. The artificial intelligence analyzing the input data provides an estimate of the mortality rate of the patient and indicates that the lifespan is full according to the situation.
The team checked the data of over 114,000 patients to test their work. According to researchers’ reports, deaths within 24 hours of admission were predictable with an accuracy of 93-95%. This shows that the Google-led research is 10% more successful than traditional forecasting methods.