Design and Ex Vivo Experimental Validations of the CMOS 256-Pixel Photovoltaic-Powered Subretinal Prosthetic Chip With Auto-Adaptive Pixels for a Wide Image Illuminance Range

IEEE Trans Biomed Eng. 2022 Jan;69(1):482-493. doi: 10.1109/TBME.2021.3098734. Epub 2021 Dec 23.

Abstract

Objective: To design and verify a CMOS 256-pixel photovoltaic-powered subretinal prosthetic chip with key advances over the state-of-the-art. The three key advances are: 1) automatic adaptation to changing background illuminance levels; 2) increase of injection charges with reduced crosstalk leakage charges, enhanced charge balance, and low process variations; 3) stable stimulation voltage to keep the safety of water window.

Methods: The novel auto-adaptive pixel circuit is designed to realize the Michealis - Menten equation (MME) so that the automatic adaptation to changing background illuminance can be achieved. Both improved biphasic constant current stimulator (CCS) via bi-directional shared electrodes (BDSEs) with optimized stimulation pattern and improved constant current generator/ring oscillator are designed to achieve the above second advance on injection charges. The closed-loop charge pump is designed to achieve the third advance.

Results: The measured dynamic range of image illuminance is increased to 54.7 dB. The maximum stimulation charge is 8.89nC. The measured stimulation current mismatch is 1.7% and the measured residual charge is 0.150 nC. The measured variations of stimulation frequencies are from 26 Hz to 29.7 Hz. The results of ex vivo tests have shown that the proposed subretinal chip can evoke spiking responses of RGCs. The function of adaptation process to background illuminance has also been verified.

Conclusion and significance: Through both electrical measurement and ex vivo tests, the functions of designed subretinal chip have been validated successfully. It is shown that the proposed subretinal chip is a promising solution for subretinal prostheses.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Electrodes
  • Prostheses and Implants*