Music streaming services such as Spotify use algorithms to recommend song lists that suit your tastes based on your preferences and those of other listeners.

Recently, a new study showed how your brain responses can be used to optimize the playing of these song lists.

Researchers in the US used machine learning algorithms to analyze the physiological responses of music listeners. They found that they could predict which songs would delight users with up to 97% accuracy.

The algorithm determines a person's mood and attention to a song based on their neurophysiological state.

If a person is more neurologically "immersed" and less "withdrawn" when listening to a song, then the song is more likely to be popular.

This is a much better predictor of a song's success than whether or not a person subjectively likes the song.

In other words, just because you consciously like a song doesn't necessarily mean other people will. However, your subconscious state may be more of a clue.

Instead of asking users if they 'like' a new song, the researchers wrote, "It is automatically assessed by wearable neurotechnology."

Even though the algorithm was only fed physiological data from listening to a song for a minute, it was able to predict popular songs with 82 percent accuracy.

This new method is more effective than previous similar studies that used brain scans to assess musical responses.

In the new experiment, participants sat in a room wearing heart sensors and listened to 24 of the latest songs played over an amplifier.

Thirteen of them were considered popular songs by the streaming platforms, but the participants didn't know which were which.

At the end of the experiment, they were asked to rank their favorite songs.

The data collected from the heart sensors was then fed into a commercial neuroscience platform that uses heart rate data to infer the state of a person's brain.

For example, oxytocin and dopamine are two neurohormones known to affect the heart.

When you sing or listen to music, there's evidence that your brainstem often releases oxytocin. In contrast, dopamine is released and interacts with the prefrontal cortex when you're particularly attentive to or "absorbed" in something.

"Using neuroscientific techniques to measure emotional responses offers artists, record producers, and streaming services a new way to delight listeners with new music," the authors write.

"Comprehensive neuroscience measures from the peripheral nervous system can accurately categorize pop and flop songs."

In theory, this information could be used to create specialized song lists for different emotional states, but it doesn't just apply to music.

"Neural prediction" could also be applied to almost any form of entertainment, allowing people to get what they want before they even realize it.

The current study is small and there are still details to work out, but as a proof of concept, it looks promising.

"Our main contribution is the methodology. Likely, this methodology can also be used to predict the popularity of many other types of entertainment, including movies and TV shows."