Can Machines Predict When We’ll Die?

Author: AB Staff

Can Machines Predict When We’ll Die?

Editor's Pick Wellness

Death is never easy, but it’s often compounded by uncertainty. We can never know which meal, conversation, hug or goodbye will be our last with our loved ones. This is especially true and even more difficult for people suffering from terminal illnesses. These people know death is coming, and soon, but they don’t know exactly when. This added stress of the unknown can greatly decrease important final moments and make planning very difficult.

Advance care planning, or planning for end of life decisions, is crucial in helping terminal patients settle what can be done ahead of time to relieve chaos, confusion and stress after death. But knowing when to begin this planning is incredibly hard for many reasons, with the main one being natural human optimism.

Families, loved ones and even doctors want to believe the best in a prognosis. While this hope is beneficial in a lot of ways, it can delay or even completely take away advance care planning, leaving patients ill-prepared for death and the bereaved unprepared for major decisions.

There may be an out-of-the-box solution to this problem: artificial intelligence. A team of researchers at Stanford Health Care is investigating the potential for artificial intelligence (AI) to help doctors better determine end-of-life timelines. This AI tool doesn’t give an exact time or date of death, but it does predict death within months or years with impressive accuracy. This data modeling has the potential to transform clinical care not only for patients and their loved ones, but also for healthcare economics.

The AI system gathers data about the patient’s medical history and millions of other patient’s medical records. It inputs this information into an algorithm to calculate the probability that the patient will die within 12 months. The AI researchers then send a report of Stanford Hospital patients with a 90% or higher chance of dying within 3 to 12 months to the necessary doctors. The system also has more specific settings and calibrations for each type of cancer for more accurate predictions.

However, it’s important to remember that while AI can help doctors, the physicians are still the ones making the decisions and analyzing the information. After all, don’t we all prefer a human touch when we’re sick or facing a loss?

Using AI to help predict end-of-life information can potentially revolutionize healthcare, but the industry leaders are proceeding with caution. There is still a lot of research and testing to be done at Stanford and many other institutions, and the doctors are hopeful yet careful to balance expectations. But who knows, maybe in a few years time we will have clear, reliable information about our own life expectancies from artificial intelligence.

This AI application is fascinating. If you would like to learn more about this and other AI possibilities, you can check out:
• How AI Can Improve End of Life Care by Scope for Stanford Medicine
• Compassionate Intelligence by Stanford Medicine
• 27 Incredible Examples of AI and Machine Learning In Practice by Forbes

Feel free to share your thoughts in the comment section below, and don’t forget to check back for new articles coming soon.