Data Scientists use Artificial Intelligence to Predict Suicide Attempts

There are many different use cases for artificial intelligence, even though most of them have yet to be explored. A Vanderbilt University data scientist has come up with a bold and radical plan to deploy AI as a way to predict suicide. That is a rather remarkable turn of events, as it could yield quite positive results. Giving others a chance to prevent people from committing suicide is invaluable, that much is evident.

Artificial Intelligence to Prevent Suicide

On paper, it makes a lot of sense to use artificial intelligence – as well as any other form of technology – to prevent suicide attempts from happening. Colin Walsh, data scientists at Vanderbilt University Medical Center, is thinking along the same lines. To be more specific, he feels AI can play a  key role in the future of predicting suicide risk and giving loved ones a chance to stop people from ending their life prematurely.

As of right now, Walsh and other scientists have successfully developed a machine-learning algorithm to predict the likelihood of people attempting suicide. As one would expect from such innovative technology, the algorithm is more than capable of accurately predicting these attempts. In fact, some people claim this algorithm is “unnervingly accurate,” which is both good and bad.

To be put this into numbers people can understand, the algorithm is between 80% and 90% accurate. It is not a bad thing to get some false positives, though, as long as it means the patient will not attempt suicide whatsoever. Failing to predict when someone would effectively attempt suicide is a factor to be a quite concerned about, though the much is evident. These results pertain to the patient’s likelihood to commit suicide in the next two years.

When reducing the timespan associated with this “investigation,” the results become a lot more accurate. More specifically, when assessing if a patient is likely to attempt suicide within the next week, the algorithm has a 92% accuracy rate. Do keep in mind all of these results are based on data widely available from hospital admissions, including patients age, gender, medications, and prior diagnoses.

So far, the team has gathered enough data from 5,617 patients to develop this algorithm. A total of 3,250 instances of suicide attempts has been recorded as a result. All of the patients in question were admitted with signs of self-harm, which is a primary indicator of future suicide attempts. Although this is still a relatively small sample size, it also goes to show the algorithm developed by the team of data scientists is definitely worth keeping an eye on.

It is evident artificial intelligence can be a valuable tool when it comes to preventing people from attempting suicide. Although this experiment is still in the early stages of development, it will be interesting to see if and whether researchers can improve upon it moving forward. Interestingly enough, a different algorithm was created to conduct similar tests looking at over 12,000 randomly selected patients with no documented history of self-harm. In this case, the algorithm was even more accurate, which is rather surprising.

Rest assured some people will feel the usage of artificial intelligence is an invasion of privacy, even if it can reduce the number of suicide attempts. There is a lot of data gathered by hospitals, which can be used for this purpose, without having to collect additional information from patients. It will be interesting to see how these algorithms evolve over time, and whether or not artificial intelligence will effectively be used to prevent suicide attempts in the future.

If you liked this article, follow us on Twitter @themerklenews and make sure to subscribe to our newsletter to receive the latest bitcoin, cryptocurrency, and technology news.