Active neural interfaces for bioelectronic medicine are envisioned to be mm-sized. Such miniaturization is at the moment hampered by the available wireless power techniques as well as the large volume the conventional hermetic packaging adds to the implant. Alternatively, conformal coatings are being explored for the protection of the implant electronics. Such approach has the potential to allow for the use of RF (radio-frequency) energy for powering, and miniaturization to the extreme of having a single IC (integrated circuit) as the whole implant (single chip implants). The longevity of conformal encapsulation can be assessed using accelerated soak tests in a dedicated apparatus in vitro, but these are usually not sufficient, as they fail to reveal additional failure modes that manifest themselves in vivo. Therefore, to investigate the performance of conformal coatings in vivo a compact, mm-sized wireless monitoring system is required. The development of such a system exhibits several challenges, mostly concerned with how to receive enough energy in such a small implant to power the monitoring sensor and transmit information regarding the integrity of the coating. In this paper we propose a system architecture for such a mm-sized wireless system, suitable for medium-to-Iong term monitoring of implants, by designing the whole system as a single monolithic IC. It is shown, by experiments, simulation or analytically that the identified challenges are possible to overcome, allowing to proceed towards the practical prototype.

@en

We present a novel method to address wavefront reconstruction problems in Adaptive Optics. The underlying estimation problem is formulated using multivariate B-splines and it is shown how the resulting constrained least-squares optimization problem can be solved using the Alternating Direction Method of Multipliers (ADMM). The proposed method is analyzed in terms of convergence and its computational benefits, compared to a centralized approach. The results indicate that due to the much more favorable scaling of the required memory and computational effort, together with good performance, the proposed solution enables solving very large scale real-time wavefront reconstruction problems that appear in next-generation adaptive optics applications.@en