Researchers from EPFL have developed a groundbreaking miniaturized brain-machine interface (BMI) capable of translating brain activity into text on tiny silicon chips. This next-generation technology, known as the Miniaturized Brain-Machine Interface (MiBMI), promises to revolutionize communication for individuals with severe motor impairments.
Brain-machine interfaces have long held the potential to restore communication and control to people with conditions such as amyotrophic lateral sclerosis (ALS) and spinal cord injuries. However, traditional BMIs have been bulky, power-hungry, and limited in practical applications. The new MiBMI, developed by EPFL’s Integrated Neurotechnologies Laboratory (INL), overcomes these challenges by offering a compact, low-power, and highly accurate solution.
Published in the IEEE Journal of Solid-State Circuits and presented at the International Solid-State Circuits Conference, the MiBMI is designed to enhance the efficiency and scalability of BMIs, paving the way for practical, fully implantable devices. Its small size and low power consumption make it ideal for clinical and real-life applications, ensuring safety and practicality.
The MiBMI is a fully integrated system, with recording and processing capabilities housed on two tiny chips with a total area of just 8mm². This innovation is part of a new class of low-power BMI devices from EPFL’s Neuro X institutes, led by Mahsa Shoaran. “MiBMI allows us to convert intricate neural activity into readable text with high accuracy and low power consumption. This advancement brings us closer to practical, implantable solutions that can significantly enhance communication abilities for individuals with severe motor impairments,” says Shoaran.
The brain-to-text process involves decoding neural signals generated when a person imagines writing letters or words. Electrodes implanted in the brain record the neural activity associated with the motor actions of handwriting. The MiBMI chipset then processes these signals in real time, translating the brain’s intended hand movements into corresponding digital text. This allows individuals with locked-in syndrome and other severe motor impairments to communicate simply by thinking about writing, with the interface converting their thoughts into readable text on a screen.
While the MiBMI chip has not yet been integrated into a fully functional BMI, it has demonstrated impressive performance in processing data from previous live recordings, such as those from the Shenoy lab at Stanford. The chip has successfully converted handwriting activity into text with an accuracy of 91%, decoding up to 31 different characters. “We are confident that we can decode up to 100 characters, but a handwriting dataset with more characters is not yet available,” says lead author Mohammed Ali Shaeri.
Current BMIs typically send data from brain-implanted electrodes to an external computer for decoding. In contrast, the MiBMI chip records and processes the information in real-time, integrating a 192-channel neural recording system with a 512-channel neural decoder. This achievement in extreme miniaturization combines expertise in integrated circuits, neural engineering, and artificial intelligence, marking a significant advance in the BMI field.
To handle the vast amount of information captured by the electrodes, the researchers adopted a novel approach to data analysis. They identified specific neural markers, dubbed Distinctive Neural Codes (DNCs), associated with each letter when a patient imagines writing it by hand. By focusing on these DNCs, which require processing only a few hundred bytes of data per letter, the system operates with remarkable speed, accuracy, and low power consumption. This breakthrough also enables faster training times, making the BMI easier to use and more accessible.
Looking ahead, Shoaran and her team are exploring various applications for the MiBMI system beyond handwriting recognition. Collaborations with other research groups at EPFL’s Neuro-X and IEM Institutes, including the laboratories of Grégoire Courtine, Silvestro Micera, Stéphanie Lacour, and David Atienza, are underway to test the system in different contexts, such as speech decoding and movement control. “Our goal is to develop a versatile BMI that can be tailored to various neurological disorders, providing a broader range of solutions for patients,” says Shoaran.
The MiBMI’s potential to revolutionize brain-machine interfaces opens up exciting possibilities for improving the quality of life for patients with severe neurological impairments.
By Impact Lab
