York researchers designing the next generation of brain-implantable devices

Digital medical illustration: Lateral (side) x-ray view (orthogonal) of human brain with blood vessels.
MohammadAli Shaeri
MohammadAli Shaeri

Researchers MohammadAli Shaeri, a post-doctoral researcher at York University, and his supervisor Amir Sodagar, an associate professor in the Lassonde School of Engineering, are developing the next generation of brain-implantable devices: high-density implants that record massive amounts of data from the brain.

Sodagar says development of brain implants is a multi-disciplinary activity, which involves cutting-edge research in microelectronic circuits, microfabrication, wireless interfacing, biological signal processing, as well as biology and neuroscience.

In a paper titled “A framework for on-implant spike sorting based on salient feature selection,” published in the June volume of the journal Nature Communications, Shaeri and Sodagar have proposed an innovative, highly accurate framework for “spike sorting” dedicated to high-density brain implants.

According to the researchers, efficient interfacing with the brain is critical to using applications such as brain-computer interfacing, neuro-prosthetic applications, and as a remedy to neural diseases, disabilities, and disorders such as blindness and Parkinson’s disease.

The ability to transmit large amounts of recorded neural data efficiently has been a challenge in designing these implants. Spike sorting, which involves isolating and assigning brain data at the level of neurons, is among the most efficient ways to reduce recorded data while preserving the main information.

Amir Sodagar
Amir Sodagar

Shaeri says that brain implants should not only be as small as possible, but also should rely on extremely low power consumption, as increases in the use of power will cause devices to heat up, potentially leading to brain tissue damage. This means that brain implant technology cannot use complex algorithms for data compression and sorting.

In their proposed spike sorting technique, which is based on the selection of the most salient features within the course of neural spikes, the researchers introduced a quantitative formulation for “feature saliency,” to best discriminate different classes of neural spike wave shapes.

Shaeri hopes the work can lead to realizing brain-implantable devices in the near future, and is interesting in the potential for wireless interfacing with the technology. He was proud to see the work published in Nature Communications.

“It’s a high-impact journal,” he explained, “it’s very hard to publish method and engineering work in such a journal. It’s a real pleasure.”

The study is available online via open access from Nature Communications.