Biomedical Engineering Reference
In-Depth Information
results obtained in these initial steps are a proof of concept which motivates further
development of the FOT device and detection algorithm.
9.5 Detecting Non-linear Distortions at Low Frequencies
Having the proof of concept available from previous sections in this chapter, the
motivation to look at lower frequencies, where viscoelasticity plays an important
role, becomes justified. However, we are facing two problems at this turn-point:
•
according to Fig.
9.4
, the mechanical device from Fig.
9.2
can only reliably excite
frequencies as low as 0.6 Hz;
•
the lower one tries to investigate the respiratory impedance, the closer one comes
to the breathing frequency of the individual under test.
Consequently, a different approach was necessary to endeavor this objective, further
detailed in the remainder of this chapter.
9.5.1 Prototype Device with Feedforward Compensation
In order to tackle the drawback of the loudspeaker (i.e. limitation in the lower fre-
quency band) and of the mechanical setup (i.e. noise, friction), a device for applying
FOT was designed by means of air-fans. The picture of the setup and its elements
are shown in Fig.
9.10
. The setup consists of two fans which are forcing the air
through a PVC tube. The fans are driven by a pulse width modulated signal gener-
ated by a PIC 18F4550 microcontroller. Pressure and flow at the mouthpiece can be
obtained using two pressure sensors and a pneumotachograph, similarly to the other
FOT devices. The excitation pressure signal is kept below a peak-to-peak variation
of 0.2 kPa at the airway opening as recommended in [
116
].
The fans create turbulences, which result in increased measurement noise. To
reduce these turbulences [
128
], the PVC tube is filled with thin tubes (i.e. cocktail
straws) of 3 mm diameter whereas the middle part of the tube is left empty to pre-
serve a good air supply for the subject. The measured pressure values are quantized
within the pressure sensors and transmitted to the microcontroller.
Two complementary compensation methods are used to suppress the linear dy-
namic behavior, the non-linear distortions generated by the measurement system and
the disturbances introduced by the breathing of the patient. Firstly, a feedforward
compensation of the excitation signal is proposed to suppress the non-linear distor-
tions and the linear dynamic behavior of the measurement system. The feedforward
compensation of the measurement system has two goals. The feedforward signal is
generated in such a way that the bandwidth of the generated pressure signal is larger
than the bandwidth of the measurement system. Additionally, the non-linear distor-
tions are suppressed by use of an iterative scheme. Secondly, the residual dynamic
