Fig. 3: Characterization of FIMOSS implant.

System diagram (a), electrical schematic (b), radio frequency communication signal (c) of FIMOSS. d Current and voltage across a variable load resistor while it sourced power from the FIMOSS implant that was wirelessly powered under the default wireless condition. The two vertical lines corresponds to the minimum (0.6 mA) and maximum (2.6 mA) current consumption of the FIMOSS implant at 3 V. e The temperature changes in vitro during the 7200 s’ continuous working with four different input voltages. (n = 4, 20 ms pulse width, 1000 ms inter-pulse interval, mean ± s.e.m.) FIMOSS implants were placed on a 1mm-thick glass slide, and the external device antennas were fixed underneath the slide. Voltages of the DC power supply were adjusted to 3.3 V, 3.6 V, 3.9 V, and 4.2 V to power the external device, which wirelessly drive the FIMOSS implants. f A representative infrared thermographic image of FIMOSS when working in vitro. (Input voltage = 3.3 V, 20 ms pulse width, 1000 ms inter-pulse interval, scale bar = 5 mm) g The temperature changes of FIMOSS implants in vivo during the 3600 s continuous working. (n = 4, input voltage = 3.3 V, 20 ms pulse width, 1000 ms inter-pulse interval) FIMOSS implants were implanted subcutaneously on the surface of pectoralis major, whose temperature were measured by an infrared thermometers h Temperature change of the µLED in vivo during a 20 s test period, where the current through the µLED was simultaneously recorded. The µLED was powered by a constant current source that was set to 2.1 mA. i The absorption coefficients and reduced scattering coefficients used for the optical simulation. j Cross-sectional view of the 3D model of a MOSD II cuffing on a typical target nerve (440 μm diameter, scale bar = 500 μm). This 3D model was used in the optical simulation, and each domain was color-coded corresponding to the color legend in (i). The 4 µLEDs were numbered, and the nerve was equally divided to eight sectors and numbered as references for the quantitative analysis result in (n). k Heat map of the light irradiance distribution in 3D optical simulation, when 1 µLED was illuminated at 2.5 mW of optical power. The color legend is the same as in Fig. 3i, m. Domains except for the activated µLED and the target were hidden. The plot was set at 40% transparency to allow visualization of the contour lines of 13 mW mm⁻² and 5 mW mm⁻² depicted in magenta and white. l, m Heat map of light irradiance in a representative 2D slice out of the 3D optical simulation result. The 2D slice was defined as the XY-plane slicing down the middle (Z-direction) of the µLED in (k). One µLED (l) was illuminated at the standard optical power of 2.5 mW. Two neighboring µLEDs (m) were illuminated at a reduced optical power of 2.0 mW. The color bar represents light irradiance in mW mm⁻². n The average irradiance in each sector of the nerve computed by optical simulation of 1-µLED (orange) and 2-µLED (yellow) illumination in Fig. 3l, m. o Time deviation (o) of 4 μLEDs’ light (blue, left y axis) and interval time (purple, right y axis) when the stimulus parameter was set to 50 ms pulse width, 1000 ms inter-pulse interval. p The correlation of pulse width and time accuracy when the pulse width was set to 10, 20, 30, 40, 50, and 200 ms. Relative (red, left y axis) and realistic time deviation (blue, right y axis) of μLEDs were recorded or calculated, respectively (n = 4, mean ± s.e.m.).