Artificial intelligence can predict when a patient dies

A team of Australian researchers developing artificial intelligence (AI) systems can predict the risk of premature death of patients by examining photographs of internal organs, according to the International Business Times. The results are published May 10 in the journal Scientific Reports.

The new AI system is a product created by researchers at the University of Adelaide, Australia. When analyzing chest images of 48 patients, this AI system predicts patients likely to die within 5 years with an accuracy of 69%.

"Accurate assessment of the biological age and life expectancy of patients is limited because doctors can not observe the inside of the body and assess the health of each organ , " said Dr. Luke. Says Oakden-Rayner, lead author of the study. "Our research focuses on 'deep learning' techniques, which allow computer systems to learn how to interpret and analyze images."

The team has not been able to accurately determine what in the picture the computer system provides for patient life expectancy, but the AI system is most successful in patients with severe chronic conditions such as emphysema ) and congestive heart failure.

According to the researchers, the AI system can be diagnosed through a large pool of data and find patterns that are difficult to detect without physicians being trained to recognize.

"Our research opens up new applications of artificial intelligence technology in medical imaging, which may provide hope for early detection of serious illness requiring special medical intervention." , Says Oakden-Rayner.

In a similar project, a team of scientists in the Google Scientists division claimed that AI technology used in the company's self-driving project could detect cancer faster and more accurately than doctors.