Science & Tech · Science & Environment · 37 mins ago
Three lab-grown brain organoids identified Braille letters with 83% accuracy
Researchers at the University of Bristol tested lab-grown brain tissue to see whether it could help process information.
The tiny organoids, made of human neurons and support cells, received electrical pulses based on Braille letters scanned by a touch sensor.
A computer program identified the letter from the organoids’ electrical activity, getting it right about 61% of the time with one organoid.
Combining signals from three organoids raised accuracy to 83%, compared with scores of 59% to 67% for each one alone.
The organoids did not learn, so the experiment shows information could be recovered from their responses, not that they can think or read independently.
Living neural tissue may offer a more energy-efficient way to compute than silicon, although this experiment does not show that such systems are ready for practical use.
Three lab-grown brain organoids identified Braille letters with 83% accuracy when their signals were combined.
A single organoid identified letters correctly about 61% of the time.
Researchers converted touch-sensor readings of Braille letters into electrical pulses received by the organoids.
The experiment used a pattern-sorting program to match the organoids’ electrical activity to letters.
The researchers said the work could help explore more energy-efficient computing, but the organoids did not learn.
- Who
- A team led by Tianyi Liu, a doctoral student at the University of Bristol, worked with Hemma Philamore and Benjamin Ward-Cherrier.
- What
- Three brain organoids identified Braille letters with 83% accuracy when their signals were combined.
- When
- Not stated.
- Where
- The organoids were grown and tested in a lab; the team was at the University of Bristol.
- Why
- The experiment explored how living neural systems might process information and potentially support more energy-efficient computing.
This story does not have two clearly opposing sides.
A sensor on a robot arm scanned 26 Braille letters, making 1,300 recordings.
The team converted touch readings into electrical pulses and sent them to the organoids.
A pattern-sorting program matched the organoids’ electrical activity to letters.
One organoid scored about 61% accuracy, while the combined signals of three scored 83%.
- Combined accuracy
- 83%
- Single-organoid accuracy
- About 61%
- Braille letters tested
- 26
- Recordings
- 1,300
- Research lead
- Tianyi Liu, University of Bristol
What people are saying
We can take information from a real tactile sensor, turn it into a form that a living neural network can receive, and then recover meaningful information from the organoid response
That gives us a practical way to connect artificial sensors with biological neural systems, which are a very interesting possible route towards more energy-efficient computing