BCIwiki (bciwiki.com) — In a participant with severe upper-limb paralysis caused by amyotrophic lateral sclerosis (ALS), a stent-electrode array implanted inside a cerebral vein recorded stronger per-channel motor modulation than a simultaneously worn scalp EEG cap, and its signal was not attenuated by the skull, although the recordings also contained prominent cardiac activity. The findings were published in the Journal of Neural Engineering on September 17, 2026.
The study was carried out by researchers from institutions including the Department of Mechanical Engineering at Carnegie Mellon University, the University of Melbourne’s medical school, Synchron, and the Rehab Neural Engineering Labs at the University of Pittsburgh. The stent-electrode array records cortical neural activity from within the superior sagittal sinus and has been tested in an early feasibility clinical trial in the United States (registration number: NCT05035823). The authors note that no direct comparison of endovascular and scalp neural recordings in humans had previously been reported.
The experiment was conducted in two sessions. The participant wore the stent-electrode array and a gel-based EEG cap simultaneously and, following visual cues, attempted motor tasks such as repeatedly flexing and extending the ankle. The team assessed signal quality by quantifying the strength of motor modulation, the ability to discriminate levels of movement effort and spatial lateralization, and used noise metrics to measure the effects of 60 Hz line noise, electrocardiographic (ECG) interference, eye blinks, jaw clenching and vocalization.
Both recording modalities showed significant modulation during attempted movement relative to rest, with the stent array often showing significantly stronger per-channel modulation in certain frequency bands and conditions. During motor imagery, modulation in both modalities was significantly weaker than during attempted movement. Spatial source localization of left versus right ankle movement did not reach statistical significance with either modality. As for artifacts, scalp EEG, lying close to superficial physiological sources, was highly susceptible to ocular artifacts and cranial muscle activity, whereas the stent array showed prominent cardiac activity.
The authors concluded that, in this participant, the larger modulation recorded by the stent-electrode array during attempted movement, its freedom from skull attenuation and its reduced artifacts in the frequency bands of interest provide preliminary evidence that the stent-electrode array can capture neural signals of high enough quality to support brain-computer interface (BCI) control. The study involved only one participant, and its results need to be validated in more patients.