Intracortical brain computer interfaces hold the potential to revolutionize neurotherapeutics, but they must overcome technological challenges such as the high data rates generated by high-channel-count neural sensors and the stringent power and volume constraints of implantable devices. In addition, the brain-wide coverage needed for a deeper understanding of brain processes challenges the synchronization between distributed neural sensors and the central neural hub. To address these challenges, we present a deterministic-latency and power-efficient serializer–deserializer (SerDes) telemetry network that effectively mitigates the synchronization issue under strict power and volume constraints. The serializer on the sensor side employs event-based sampling and a packet-based address-event representation transmission protocol, achieving a low power consumption of only 127 µW and a low latency variation <10 µs. A crystal-free clock source is employed on the sensor side to minimize power consumption, with serialized data encoded using Manchester coding scheme. The deserializer on the hub handles the bit period uncertainty by counting and extracting the bit period of received data with a clock only ∼2.2× faster than the serializer clock. The proposed counting-based Manchester decoder achieves a wide frequency coverage up to 204 000 ppm of frequency variation. The deserializer achieves a measured Manchester decoding bit error rate (BER <10−6), with a total estimated power consumption below 415 µW. The SerDes performance has been validated with in vivo pre-recorded data, demonstrating a compression ratio greater than 7, while preserving a high signal fidelity with an average RMSE <6 µVRMS.

An ultrasound powering TX ASIC for brain implants with autonomous on-chip standing-wave peak tracking for RX power (PDL) regulation is presented. With a proposed adiabatic power-sensing scheme, the TX consumes 43μW for power tracking, and the system achieves a settling time of <90ms while using the standing-wave peak-tracking FSM. The TX can achieve PDL improvement of up to 2.7× with a power-tracking accuracy of 82%.

A near-field galvanic coupled transdural telemetry ASICs for intracortical brain-computer interfaces is presented. The proposed design features a two channels transmitter and three channels receiver (2TX-3RX) topology, which introduces spatial diversity to effectively mitigate misalignments (both lateral and rotational) between the brain and the skull and recovers the path loss by 13 dB when the RX is in the worst-case blind spot. This spatial diversity also allows the presented telemetry to support the spatial division multiplexing required for a high-capacity multi-implant distributed network. It achieves a signal-to-interference ratio of 12 dB, even with the adjacent interference node placed only 8 mm away from the desired link. While consuming only 0.33 mW for each channel, the presented RX achieves a wide bandwidth of 360 MHz and a low input referred noise of 13.21 nV/√ H z. The presented telemetry achieves a 270 Mbps data rate with a BER < 10 −6 and an energy efficiency of 3.4 pJ/b and 3.7 pJ/b, respectively. The core footprint of the TX and RX modules is only 100 and 52 mm 2 , respectively, minimizing the invasiveness of the surgery. The proposed transdural telemetry system has been characterized ex-vivo with a 7-mm thick porcine tissue.

State-of-the-art intracortical neural recording and stimulation systems rely on subdural implants tethered to a cranial implant which itself has a wireless power and data link to the outside world [1] (Fig. 6.2.1). However, this tethered configuration poses challenges such as scarring and potential damage to the surrounding tissue due to strain and micromotions, making this approach unsuitable for chronical implants [2]. Consequently, there is growing interest in wireless connections between cranial and subdural implants. This paper focuses on wireless powering between implants, traversing the dura and cerebrospinal fluid (CSF) tissue layers over distances of 0.5 to 1cm (transdural powering). With modern burr-hole craniotomy, the hole drilled in the skull is 6mm in diameter, limiting the available size for the TX. Moreover, the power dissipation of the TX must be low to keep tissue heating below 1°C [3]. RF and optical modalities suffer from higher attenuation in tissue compared to ultrasound (0.6dB/cm/MHz) [4]. Furthermore, for transdural powering, power losses from reflections at medium interfaces (e.g., skull) are avoided, making ultrasound (US) a prime candidate for efficient in-body wireless power transfer. US is also preferable to inductive powering since US beam steering up to large angles (>45°) is needed to maximize power delivery and compensate for brain micromotions of up to ±4mm [5] and misalignment during surgery. However, prior art US driving systems either use single-phase transducer driving [6, 7], incapable of beam steering, or use class D drivers with low power transfer efficiency (PTE) [8, 9]. A phased array with increased driving efficiency was presented in [10], but it cannot perform beam steering without grating lobes that can be eliminated with miniature transducers with a pitch close to λ/2. To facilitate direct integration between CMOS and the transducer array, the CMOS driving units should also be pitch matched [8, 9].

Designing processors for implantable closed-loop neuromodulation systems presents a formidable challenge owing to the constrained operational environment, which requires low latency and high energy efficacy. Previous benchmarks have provided limited insights into power consumption and latency. However, this study introduces algorithmic metrics that capture the potential and limitations of neural decoders for closed-loop intra-cortical brain-computer interfaces in the context of energy and hardware constraints. This study benchmarks common decoding methods for predicting a primate’s finger kinematics from the motor cortex and explores their suitability for low latency and high energy efficient neural decoding. The study found that ANN-based decoders provide superior decoding accuracy, requiring high latency and many operations to effectively decode neural signals. Spiking neural networks (SNNs) have emerged as a solution, bridging this gap by achieving competitive decoding performance within sub-10 ms while utilizing a fraction of computational resources. These distinctive advantages of neuromorphic SNNs make them highly suitable for the challenging closed-loop neural modulation environment. Their capacity to balance decoding accuracy and operational efficiency offers immense potential in reshaping the landscape of neural decoders, fostering greater understanding, and opening new frontiers in closed-loop intra-cortical human-machine interaction.