A range of new applications of neural interfaces have developed in the past few decades. Cochlear implants led to retinal implants, motor cortex recording prostheses, sensory prostheses, and closed-loop sense and control devices that have yet to be developed. As these new systems are invented, the number of recording or stimulation channels has increased, and our ambition has outpaced certain areas of technology, specifically the hermetic packaging of implantable devices. For a device implanted into the body for many years, even decades, electronic circuits must be protected from bodily fluids. Polymer coatings are sufficient to prevent liquid water from reaching the circuits for many months, but water vapor can penetrate the loosely cross-linked chains and condense on the electronics. Therefore, crystal structures are used in implanted devices to block water vapor. These packages are often made of titanium and/or ceramic materials. Existing hermetic packages are filled with helium before being sealed, and are tested for helium leakage. This gives an easy translation to water vapor leakage into the device. This process allows every single device to be tested, and a prediction made for its useful lifetime inside the body. However, the technology to develop titanium or ceramic packages is limited in its ability to create high density feedthroughs. My group has worked with vendors to create a feedthrough with a density exceeding 300 channels per cm2, but the manufacturing and assembly process has been difficult. Many devices would benefit from a density closer to 1000 channels per cm2, requiring the development of new packaging technology. We propose to use the same semiconductor technology used to create microchips to build the package around the chip. This allows us to easily register to the chip pads to create connections, reducing assembly complexity. We will add alternating layers of polymer and crystalline materials, creating a hermetic barrier that is mechanically sound. However, such a device has no internal space to accommodate helium for the traditional lifetime projection test. We have developed procedures and circuits to measure very high impedances, and we propose to embed interdigitated electrodes into the package layers, measuring the impedance to detect leakage. We will measure initial water penetration to predict package lifetime, and we will also create a system to monitor the impedance over the lifetime of a device, allowing for warnings as the device nears the end of its lifetime. We feel that these advances in high-density packaging and device testability will enable higher- density versions of the VA peripheral nerve stimulator, as well as other existing implantable devices, and inspire new devices and applications that will help our veteran population.
The VA is developing a high-density peripheral nerve stimulator, as well as other types of implantable neurostimulators, to restore functional limb movement or sensation in veterans with spinal cord injuries or limb injury or loss. The proposed effort is aimed at developing next-generation packages to produce implantable medical devices that are smaller and more easily implanted, and have larger numbers of stimulation or recording channels. Specifically, we will explore microfabricated hermetic, or nearly hermetic, coatings for medical implant packages, as well as circuits to sense moisture penetration into the package layers that could cause failure of the device. These new technologies will enable smaller, more functional neurostimulation devices to serve the VA patient population.
