Long-Term Memory Neural Circuits, Fast and Precise - Brain Theory from a Circuits and Systems Perspective - page 139

07 2 cm 2 P LOSS ¼ 0 : 0 : 0003 ¼ 1 : 47 W = :

μ

Energy in joules is the integration of power over time. A fairly common type of
neural pulse can be modeled as equivalent to being high for 1 ms and low for 3 ms,
that is, high for ¼ of the time and low for of the time. Voltages in a pulse train go

¾

from about 70 to +40 mV. So power consumption in Watts during a pulse burst

may be calculated as

3
4 P LOSS þ

1
4 0

04 2

P ACT ¼ : 0 : 0003

3
4 1

1
4 0

¼ : 47 þ : 48

cm 2

¼ 1 : 222 W =

μ

So membrane power actually drops during a pulse burst. A decrease in power
means a decrease in oxygen consumption, hence a brighter spot in an fMRI, all else
being equal. The contrast is 1.47 W/cm 2 or 16.87 % of resting

power density, about what is observed. This could be a simple explanation of why
during heightened neural activity, there actually occurs an increase in the magnetic
resonance signal.

1.22 0.25

μ

Alternative Multiwrite Using Toggle Elements

Each blank word may be equipped with a Long-Term memory Toggle all of which
are initially set to false or rest. But the first toggle for the first blank word must have
its toggle set true for a brief time. Figure 8.11 shows a circuit that assumes a signal

Fig. 8.11 Multiwrite using toggle elements