Connexus is the BCI system developed by Paradromics, a Texas-based neurotechnology company. The system uses a novel microwire electrode array architecture — thousands of fine metal wires bundled together — to achieve high channel counts for intracortical recording. Paradromics positions Connexus as a high-bandwidth data interface for the brain, targeting speech and motor BCI applications.
Architecture
The Connexus system consists of:
- Microwire bundle array: Thousands of fine platinum-iridium microwires (approximately 20 micrometers diameter each) bundled together and cut to create a planar recording surface. Each wire tip serves as an individual electrode.
- Hermetically sealed implant: A biocompatible housing containing signal processing electronics, wireless data transmission, and power management
- Wireless data link: High-bandwidth wireless connection to an external processing unit
- External processor: Decoder hardware and software for real-time neural signal processing
Microwire Approach
Paradromics' microwire bundle approach differs from both the Utah Array (rigid silicon shanks) and Neuralink (flexible polymer threads):
- Scalability: Microwire bundles can be scaled to very high channel counts (1,000-10,000+) by adding more wires to the bundle
- Penetrating recording: Like the Utah Array, microwires penetrate cortical tissue for intracortical recording of single units and multi-unit activity
- Manufacturing: Wire drawing and bundling leverage established metallurgical processes, potentially offering manufacturing advantages over lithographic fabrication
Clinical Development
Paradromics has received FDA Breakthrough Device Designation for Connexus. The company has conducted chronic animal studies demonstrating long-term recording stability with the microwire architecture. As of early 2026, Paradromics is pursuing IDE authorization for first-in-human studies.
Target Applications
Paradromics emphasizes high-bandwidth data throughput as the key differentiator for Connexus, targeting applications that require decoding complex neural signals — particularly speech BCIs, where high channel counts and spatial resolution can improve phoneme and articulatory decoding accuracy.