What If An Algorithm Could Help Prevent Suicide?


Depression and suicide are not things that should be taken lightly. In fact, according to a recent newspaper article, the rate of people committing suicide is on the rise in Sri Lanka. One of the main issues with this is that we cannot exactly identify those suffering from mental illnesses or diagnose their mental health condition. But that may all change very soon.

Scientists have trained a computer program to identify people with suicidal thoughts, based on their brain scans. Though the study is relatively small, according to the researchers, the method developed could one day be used to diagnose the mental health of individuals.

Suicide is an uncomfortable topic for people

Almost One Million people fall victim to depression and suicide each year and predicting suicide remains to be a difficult challenge. The main reason for this is that a majority of people prefer not to talk about it or feel uncomfortable talking about depression and other mental health problems.

Image Credits: Medium

According to a study published in the Nature Communications journal, researchers observed the brain activity of 34 adults divided into two groups. 17 had suicidal thoughts whereas the other 17 didn’t. They read out positive and negative words and also were asked to think about words such as “evil” or “praise.” While this was happening, they underwent a brain scan called fMRI. The gathered data was then sent to an algorithm that learned to predict who had suicidal thoughts with almost 91% accuracy. In addition, the algorithm also successfully predicted whether someone had attempted suicide before with 94% accuracy.

It’s not perfect, but it’s a start

It’s not a perfect algorithm. But it’s a start. While medical tests provide more conclusive results, brain scans are rather expensive. The new algorithm would be a welcome addition to understand and prevent depression and suicide. When we think about something, the neurons in our brain act a certain way. One word would have a different neuron pattern when compared to another word. Analyzing these patterns would be more accurate than other brain studies that only look at the general brain region that is activated.

According to Blake Richards, who is a neuroscientist at the University of Toronto, the results are curious, but might not be strong enough be used as a regular diagnosis. While the accuracy of the results may be high, in order for the program to be useful in a clinical setting, and to justify any type of medical intervention, the algorithm and the results it provides would have to be almost if not definitely perfect.

If perfected, this would indeed be a very helpful tool for those suffering from depression. No longer would they have to dwell on these conditions and we can even save their lives and understand why and how depression works, and prevent suicide, building more ways to lead a happier life.


Please enter your comment!
Please enter your name here