Australian scientists have built an artificial intelligence system that can predict whether or not you will die soon by looking at images of your organs.
Researchers at the University of Adelaide used AI to analyse the medical imaging of 48 patients' chests. The system was able to predict which patients would die over the next five years with about 69% accuracy. Based on deep learning, the AI draws conclusions based on large repositories of data previously fed into it.
"Although for this study only a small sample of patients was used, our research suggests that the computer has learned to recognise the complex imaging appearances of diseases, something that requires extensive training for human experts," said lead author Dr Luke Oakden-Rayner.
The AI could be crucial to future medical machine learning software as most current systems can only diagnose diseases and not predict death. Also, examining an individual organ can be time-consuming; and AIs such as the one used by the researchers can determine the health of an organ much faster, helping doctors create detailed treatment plans.
"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," said Oakden-Rayner.
At the London Institute of Medical Services, a similar AI was recently developed that examined hearts and claimed that it could predict whether or not a patient would die within the next year with 80% accuracy compared to 60% accuracy by an average human doctor.