Page 140 - Biomedical Engineering and Design Handbook Volume 2, Applications

DESIGN OF RESPIRATORY DEVICES 119

spirometer. In this device, the gas to be measured is contained within a bellows whose expansion is
recorded via a pen or a rotational potentiometer.
Volume-displacement spirometers offer the advantage of simple construction and use. They do
not require computers or processors for simple volume and time measurements. Perhaps most
importantly, they are easy to calibrate and do not depend on the composition of gases they are used
to measure. However, they do suffer from some disadvantages. First, they are bulky. Water-sealed
spirometers, in particular, can be heavy when they are filled with water, and they are prone to spillage
when tipped. Second, owing to their mechanical action, they have a limited frequency response and
are not well suited to rapidly changing signals (although they do have satisfactory frequency
response for most common measurements). Last, the maximum volume they can measure is limited
by their size. Thus, for an experiment in which tidal volume is measured over a period of 5 minutes,
the volume-displacement spirometer would be difficult to employ, without a series of complicated
valves, as it would be filled before the experiment was over. Nevertheless, the volume-displacement
spirometer remains popular for simple measurements of volume and time.
A completely different approach, applicable only to volumes of gas as inspired or expired from the
body, relies on the changes in chest (and sometimes abdominal) geometry that accompany breathing.
One design uses two elastic bands, one placed around the chest, the other around the abdomen. The
bands contain strain gages that measure the relative expansion of each of the compartments. This device
requires calibration with each patient or subject on whom it is to be used. It is affected by movement
artifact, changes in patient position, or changes in the relative movements of chest and abdomen dur-
ing breathing. It is best employed on patients at rest, during quiet breathing, such as during sleep. Its
accuracy seldom exceeds ±20 percent. A similar device uses electrodes placed on the chest wall to
measure transthoracic impedance as a means to estimate changes in chest geometry. These devices
offer the advantage of easy long-term measurements without the need for mouthpieces or other cum-
bersome connections, but their relative inaccuracy limits their usefulness.

4.4.2 Flow
Flow is the time-derivative of volume, or

V = dV (4.8)
dt
Thus, any device capable of measuring either volume or flow can also report the other, given an
appropriate time measurement and the needed processing. For this reason, and others given below,
flow-measuring devices have become popular methods for measuring both volumes and flows
(although volume-displacement spirometers also can easily calculate and report flow measurements).
There are three common groups of flow-measuring sensors. The pressure-drop pneumota-
chometers rely on the Poiseuille equation by imposing a fixed resistance on the gas flow and mea-
suring the pressure drop across the resistance. Assuming laminar flow, the pressure drop will be
linearly related to the flow rate. The resistive element may come in several forms, but two of the
most common are the Lilly and the Fleisch (Fig. 4.5). The Lilly type uses a fine screen (similar to
a window screen) to provide resistance. Often, three screens are used, with the outer screens meant
to maintain laminar flow and to prevent debris and exhaled droplets from damaging the middle
screen across which the pressure drop is measured. The Fleisch type uses a group of small parallel
tubes as its resistive element.
In their simplest designs, both types rely on laminar flow to maintain a linear output. Thus, they
may be rated for a maximum flow rate, above which linearity is compromised, as predicted by
increases in the Reynold’s number. In practice, deviations from linearity may be seen even below the
maximum rated flow. Sometimes this may be due to suboptimal geometry between the device and
the connecting tubes and valves. Another cause of inaccuracy may be condensation of water from
the warm moist exhaled gas as it contacts the colder surfaces of the pneumotachometer. In addition
to causing computational difficulties because of changes in gas volume as water vapor is lost, the
condensed water that is deposited on the resistive element may change its resistance and diminish its