.png){kind=icon}
.png){kind=icon}
Return
Gain Modulation Using Shunting Inhibition in a Biased Subthreshold CMOS Dendrite
.png){kind=icon}
PRE
AI
L
S
F
S
DOI：10.1109/OJCAS.2026.3667082.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
The study of computational properties in biological nervous systems serves two important purposes: it examines the physical mechanisms underlying biological computation while also providing new models that can be utilized in bio-inspired artificial systems. We analyze a dendrite circuit designed to model and implement shunting inhibition, a mechanism in which the dominant effect is a change in transmembrane conductance that produces multiplicative or divisive effects in the membrane potential's response when integrating synaptic inputs. Building upon our previous work, this analysis focuses on the subthreshold-linear operating region of the dendrite circuit while also showing the effects of circuit biasing, which introduces the ability for approximating linear equations with an intercept term rather than performing multiplication alone. The real-time interaction between shunting conductance and biologically realistic current input is also studied, demonstrating the multiplicative effect of shunting inhibition using time-varying inputs of different modalities. The dendrite circuit is implemented on an analog VLSI platform and uses subthreshold transistors to emulate the electrical properties of membrane channels. This implementation serves as an analog computational unit for neuromorphic system design, exemplifying how biophysical mechanisms in dendrites can inform analog neuromorphic computation.
Keywords:
Dendrites (neurons)
Integrated circuit modeling
Transistors
Field programmable analog arrays
Electric potential
Voltage
Biomembranes
Biological system modeling
MOSFET
Neuroscience
Bio-inspired computing
CMOS analog integrated circuits
dendrites (neurons)
field programmable analog arrays
neuromorphic engineering
very large scale integration
Journal
I
IF:
2.4
Papers:
4.5K
Citations:
387
Organization
Cited Papers