Scientists Are One Step Closer to Predicting the Human Lifespan Using Artificial Intelligence
Scientists are closer to predicting patients' lifespans using medical imaging analysis.
This intriguing claim was made by a research team at the University of Adelaide in a first-of-its-kind study published in Scientific Reports. The new technique could also allow for early prediction of serious illnesses and early medical intervention.
Researchers from the University's School of Public Health and School of Computer Science used artificial intelligence to analyze medical images of 48 patients' chests. The computer image analysis was able to predict which patients would die within 5 years with 69% accuracy, as compared to 'manual' predictions by clinicians, who are hindered by the inability to go inside the body of their patients and measure the health of each organ in order to predict longevity.
"Predicting the future of a patient is useful because it may enable doctors to tailor treatments to the individual," says lead author Dr. Luke Oakden-Rayner, a radiologist and PhD student with the University of Adelaide's School of Public Health, on ScienceDaily.com. It may be especially useful for serious chronic diseases such as emphysema and congestive heart failure.
Dr. Oakden-Rayner said that the research suggests that the computers have started to recognize the "complex imaging appearances of diseases" through a process called deep learning, which would require extensive training for human researchers. Specifically, the computers collect large amounts of data and detect subtle patterns which may indicate disease.
This is significant because it may open the door for new ways that computer image analysis and medical imaging can aid in the early detection of illness, as well as serious conditions such as the onset of a heart attack. As explained in the study's abstract:
"This work demonstrates that radiomics techniques can be used to extract biomarkers relevant to one of the most widely used outcomes in epidemiological and clinical research -- mortality, and that deep learning with convolutional neural networks can be usefully applied to radiomics research."