In a significant step forward for assistive technology, researchers at Carnegie Mellon University have unveiled a new non-invasive brain-computer interface (BCI) that enables individuals to control robotic fingers solely with their thoughts.
For years, brain-computer interfaces have offered hope to those living with motor impairments, but many systems have relied on invasive techniques requiring brain surgery. While these approaches offer high precision, their risks and maintenance needs have kept them out of reach for most people. Now, a team led by Professor Bin He, a pioneer in the field, has developed a groundbreaking alternative.
Using electroencephalography (EEG), a non-invasive method for reading brain signals, the researchers developed a system that decodes a person’s motor intentions and translates them into real-time finger movements on a robotic hand. Through the power of deep learning, participants in the study were able to perform coordinated two- and three-finger tasks by simply imagining the movement. They didn’t have to lift a finger literally.
“In our study, subjects were able to successfully perform multi-finger control tasks just by thinking about it,” said Bin He, professor of biomedical engineering at Carnegie Mellon University.
The success lies in a novel neural network developed by the team, which refines and continuously decodes brain activity into precise, multi-finger actions. This deep-learning framework overcomes the spatial limitations that typically hinder noninvasive EEG systems.
“This was accomplished with the assistance of a novel deep-learning decoding strategy and a network fine-tuning mechanism for continuous decoding from noninvasive EEG signals,” the study explains.
This achievement is more than an academic milestone; it’s a step toward transforming rehabilitation, assistive devices, and even human-computer interaction.
Professor He, whose lab has previously demonstrated mind-controlled drones and robotic limbs, emphasized the broader impact: “Improving hand function is a top priority for both impaired and able-bodied individuals, as even small gains can meaningfully enhance ability and quality of life.”
Unlike surgically implanted BCIs, this system is risk-free, fully external, and adaptable to diverse environments, making it accessible to a wider range of users, including those recovering from strokes or living with paralysis.
“The insights gained from this study hold immense potential to elevate the clinical relevance of noninvasive BCIs and enable applications across a broader population,” He added.
Beyond medical rehabilitation, this level of control could revolutionize how humans interact with machines, enabling tasks like typing, handling small tools, or gaming, using only brainpower. It also moves us closer to developing prosthetics that mimic the natural dexterity of the human hand, long a goal in the field.
The research was published in Nature Communications.