This computer can determine when you’ll die
This study is the first of its kind
Research from the University of Adelaide is moving scientists one step closer to predicting patients’ lifespans by analyzing images of their organs.
Artificial intelligence is being used to analyze medical imaging of patients’ chests, which can help predict which patients would die in the next five years. The predictions were just under 70 percent correct.
Dr. Luke Oakden-Rayner, a radiologist and PhD candidate at the University of Adelaide’s School of Public Health, told Science Daily that there were definite benefits to this advancement in technology.
“Predicting the future of a patient is useful because it may enable doctors to tailor treatments to the individual,” he said in an interview with Science Daily.
The technology is significant because doctors were unable to provide people more accurate lifespan information since they couldn’t look inside patients’ bodies and measure each organ’s health.
The research investigated “deep learning,” a method where computers learn how to understand and analyze different images.
The study only utilized a small sample of patients, however, the research indicates that the computer learned to analyze and recognize different images of diseases or other health-related issues in organs. This practice would otherwise require years of training for an expert, according to Oakden-Rayner.
Some of the predictions made in the sample were for severe chronic diseases including congestive heart failure and emphysema, according to Science Daily.
“Instead of focusing on diagnosing diseases, the automated systems can predict medical outcomes in a way that doctors are not trained to do, by incorporating large volumes of data and detecting subtle patterns,” Oakden-Rayner said to Science Daily.