The extremely thin BISC curves to the shape of the patient’s brain. (Image Credit: Columbia University)
Researchers from Columbia University and other institutions developed the Biological Interface System to Cortex (BISC), a paper-thin, wireless miniaturized brain implant with high data throughput. The BISC enhances human-computer interaction and has potential applications in the healthcare world, assisting with the treatment of neurological conditions, including ALS, spinal cord injury, epilepsy, stroke, and blindness.
The BISC is built on a silicon chip and includes a wearable relay station along with dedicated system software to transmit large amounts of data from the brain to an external computer. With a volume of 3mm³ and a thickness of 50 micrometers, the implant is made of a complementary metal-oxide-semiconductor (CMOS) integrated circuit. Its high flexibility allows it to conform to the surface of the brain. The micro-electrocorticography device incorporates 65,536 electrodes, supports 1,024 channels for neural recording, and 16,384 channels for electrical stimulation. Manufactured using standard semiconductor fabrication techniques, this implant can be produced on a large scale using modern industrial processes.
The chip includes a wireless power circuit, a radio transceiver, digital control logic, data converters, and analog components for recording and stimulation, along with power management. It also has a battery-powered external relay station that supplies power to the implant and manages two-way data transfer via an ultrawideband radio link with data rates up to 100 Mbps. To external systems, the relay appears as an 802.11 WiFi device that communicates data between the brain and computer.
Instead of using computing frameworks, BISC depends on a custom instruction set and a comprehensive software ecosystem tailored for brain-computer interfaces. During experiments, the system transmitted detailed neural signals into advanced machine learning and deep learning models that interpret cognitive signals like intention, perception, and internal neural states.
The team is now working on making the BCIS available for clinical use. So far, they have established and optimized surgical techniques to implant this ultrathin device in preclinical models. The trials confirmed the chip reliably captures neural activity and maintains stable performance over long periods. Building on those results, the team is now focusing on human trials for short-term neural recordings during surgical procedures.
Additionally, the team is accelerating the transition from research to real-world applications with the launch of Kampto Neurotech. This startup is manufacturing commercial-grade versions of the chip for preclinical research and is seeking investments to push the technology to clinical use in humans.
Have a story tip? Message me here at element14.