An abundant amount of new research continues to investigate how computer algorithms and artificial intelligence—computer learning—could become more useful to us in the medical industry. For one, a new study reported this week that deep learning algorithms can be quite accurate in terms of predicting the onset of Alzheimer’s disease as far as 6 years in advance.
Essentially, the study claims, computers can use “training dataset” deep learning algorithms to “teach themselves” to accurately predict if and when a specific event would be most likely to happen.
Lead study author Dr. Stephen Weng comments, “Preventative healthcare is a growing priority in the fight against serious diseases so we have been working for a number of years to improve the accuracy of computerized health risk assessment in the general population.”
For the study, researchers collected health data from about half a million people between the ages of 40 and 69. These participants had voluntarily participated in the UK Biobank study, which collected this data between 2006 and 2010, following them until 2016. With this data, then, the research team developed a new system of deep learning algorithms to predict premature death from chronic disease. Also using death records from the Office of National Statistics, the United Kingdom cancer registry, and “hospital episodes” statistics, Dr. Weng explains how the team was able to map out the predictions of mortality data.
The University of Nottingham (UK) assistant professor of epidemiology goes on to say, “Most applications focus on a single disease area but predicting death due to several different disease outcomes is highly complex, especially given environmental and individual factors that may affect them.”
Dr. Weng notes that this is a “major step forward” in terms of developing a new, holistic approach to detecting premature death using machine learning. He adds, “This uses computers to build new risk prediction models that take into account a wide range of demographic, biometric, clinical, and lifestyle factors for each individual assessed, even their dietary consumption of fruit, vegetables, and meat per day.”
Of course, this is only the beginning of what this technology can do. Over the next few years we will likely see more advancement in this realm, particularly in improving the ways to communicate what we learn in more concise ways to benefit not just doctors but patients as well.
The results of this study have been published in the journal PLOS ONE.