| Contents |
| A major problem in neuroscience is to unveil brain's connectivity. A local simplified situation is finding out which signal is receiving a single neuron subjected to a bombardment of synaptic inputs and discern the temporal contributions of excitation from those of inhibition. This quantitative information is important for the integrative properties of cortical neurons. Due to the multitude of synaptic contacts, obtaining direct measurements of the synaptic input becomes unrealistic. Therefore, inverse methods appear as an alternative to estimate the input conductances from experimental measurements. An extended strategy is filtering the voltage and then assume an input-output relationship. We have shown, using computational models, that this linearity hypothesis is unreliable both during spiking and in non-spiking regimes with ionic activity. We will lecture on the latest situation using a conductance-based model endowed both with an afterhyperpolarizing and low-subthreshold currents to show that the subthreshold activity can lead to significant errors in synaptic conductance estimation. Our results add a warning message about extracting conductance traces from intracellular recordings and the conclusions concerning neuronal activity that can be drawn from them. They also stimulate challenging questions in developing theoretical efficient methods, specially as an inverse problem in dynamical systems. |
|