People with spinal cord injuries often lose the ability to move their arms or legs. In such cases, the nerves in the limbs usually remain healthy, and the brain continues to function normally. The problem arises because damage to the spinal cord blocks the transmission of signals between the brain and the body, resulting in loss of movement.

This disruption has motivated scientists to search for ways to restore communication between the brain and the body without needing to repair the spinal cord itself. In a study published in the journal APL Bioengineering, specialists from universities in Italy and Switzerland explored whether electroencephalography (EEG) could help bridge this gap. Their goal was to determine if EEG could detect brain signals related to movement and potentially transmit them back to the body. Even when a person tries to move a paralyzed limb, the brain still generates electrical activity corresponding to that action. If these signals can be detected and accurately interpreted, they could be sent to a spinal cord stimulator that activates the nerves responsible for limb movement.

Previously, most similar studies relied on surgically implanted electrodes that record motor signals directly from the brain. While such systems have shown promising results, the research team wanted to find out if EEG could offer a safer alternative. EEG systems are essentially caps with electrodes that record brain activity from the surface of the head. Despite certain challenges in their use, this method avoids the risks associated with placing devices inside the brain or spinal cord.

However, using EEG to decode movement attempts presents significant challenges for current technology. Since EEG electrodes are placed on the scalp, it is difficult for them to pick up signals from deep brain regions. This issue is less pronounced for hand and arm movements, but signals controlling leg and foot movements are harder to detect because they originate from areas closer to the center of the brain.

To analyze EEG data more accurately, the researchers used a machine learning algorithm designed to work with small and complex datasets. During testing, patients wore EEG caps and attempted to perform simple movements. The team recorded brain activity and trained the algorithm to distinguish signals at different moments in time.

The system was able to successfully determine when patients were trying to move and when they remained still. However, it struggled to differentiate between attempts at different types of movements.

The scientists believe their method can be improved. They plan to refine the algorithm so it can recognize specific actions, such as standing up, walking, or climbing stairs. The team also hopes to study how the decoded signals can be used to activate implanted stimulators in patients recovering from spinal cord injuries.

If this approach proves successful, non-invasive brain scanning could move closer to helping people regain movement after paralysis.