EEG technology could be a useful tool for linking brain signals with limb movements

People with spinal cord injuries often lose some or all function in their limbs. In most patients, the nerves in the limbs function normally and the neurons in the brain still function, but the damage to the spinal cord prevents the two areas from communicating.

In APL Bioengineering, published by AIP Publishing, researchers from Italian and Swiss universities conducted an initial feasibility study to explore whether electroencephalography (EEG) could be a useful tool for linking brain signals to limb movements.

When a patient attempts to move a paralyzed limb, the brain generates a series of signals that correspond to the movement. If these signals could be read and decoded, they could be relayed to a spinal cord stimulator to control the nerve endings in that limb.

Previous research has focused on implantable electrodes for reading movement signals. Although this approach met with some success, the authors wanted to explore the potential of EEG technology.

EEG devices typically look like a cap filled with electrodes that measure brain activity. And while a nest of wires may seem intimidating, the authors say this approach is preferable to implanting devices in the brain or spinal column.

It may cause infection. It’s a separate surgical procedure. We were wondering if we could avoid it. ”

Laura Toni, Author

But using EEG to decipher attempted limb movements is pushing the limits of the technology. The electrodes are placed on the surface of the patient’s head, making it difficult to pick up signals generated in deeper regions of the brain. This is only a minor obstacle when applied to arm and hand movements, but becomes more difficult when applied to legs and feet.

“The brain controls lower limb movements primarily in the central region, while upper limb movements are controlled more laterally,” says Toni. “It’s easier to spatially map what you’re trying to decipher compared to the lower extremities.”

To aid in deciphering EEG signals, the authors employed machine learning algorithms designed to sift through this type of limited dataset. In the test, the researchers fitted patients with an EEG monitor and asked them to perform a series of simple movements. They collected the resulting data and used an algorithm to classify the range of possible signals.

They found that although they were able to detect the difference between attempted and non-attempted movements, they had trouble distinguishing between specific signals.

The researchers have ideas for how to increase the effectiveness of their approach in future studies. They hope to improve the algorithm to recognize attempts at various movements, such as standing, walking, and climbing, and explore ways to use that data to trigger those movements on implants in recovering patients.