Since the 1960s, it is held that when a neuron fires, a nerve spike passes only through the selective branches, the calculated choice is a key to learning by rewiring. It is argued by chemically estimating the membrane's ion channel density that different axonal branches get active to pass the spike-branches blink at firing at different time domains. Here, using a new time-lapse dielectric imaging, we visualize the classic branch selection process; thenceforth, hidden circuits operating at different time domains become visible. The fractal grid of coaxial probes captures wireless snapshots of material's vibration at various depths below the membrane by setting a suitable frequency. Thus far, branch selection observed emitted energy or particle but never the emitters, what they do. As each dielectric material transmits and reflects signals of different frequencies, we image live how filaments search for many branch-made circuits, choose a unique pathway 103 times faster than a single nerve spike. It reveals that neural branches and circuit visible in a microscope are not absolute, there coexist many circuits each operating in different dime domains, operating at a time.NEW & NOTEWORTHY Using dielectric resonance scanner, we show electromagnetic field connections between physically separated neurons. Electromagnetic field creates field lines that pass through gap junctions, connect Axon initial segment with the dendrites through Soma, and connect axonal or dendritic branches even if there is no synaptic junction. Consequently, many distinct loops connecting various branches form coexisting circuits. Our discovery suggests that physically appearing neural circuit is a fractional view of many simultaneously operating circuits in different time domains in a neural network.
Keywords: dielectric resonance imaging; hippocampal neuron cell; neural network; neuron spike.