For monitoring the respiratory rate of the bedridden or bedbound
patients/elderly, we need to consider the different
sleeping positions that will affect the sensors. According to
a famous study by Professor Chris Idzikowski, director of
the Sleep Assessment and Advisory Service in United
Kingdom, there are six common sleeping positions as
shown in Fig. 11.
We performed many experiments using many different
users repetitively on the smart FBG pressure sensor system
for monitoring the respiratory rate of a person on the bed
according to the six sleeping positions mentioned above. It
was found that most of the respiratory rates of patients are
around 10 to 25, which is the normal respiratory rate for
adults. In order to validate the accuracy of the respiratory
rate from our system, we asked the users to mentally count
the number of breathing without looking at the system, and
almost all values given by users tally with the ones given
by the system. We are still in the process of benchmarking
it using the gold standards which are the respiratory
inductive plethysmogram (RIP) and the airflow measurement
methods. Through trials conducted in the laboratory
with a sample group of 10 subjects, the system showed
maximum error of 1 breaths per minute as compared to
manual counting.
We also did some preliminary study to measure heart
rate relying on simple fast Fourier transform (FFT)
techniques to measure the rate after low-pass and bandpass
filtering of the signal, but the sensor was placed on top
of the mattress in this case. The study showed that the heart
rate detected from the FBG sensor is very close to that
measured by off-the-shelf pulse oximeter (brand: NONIN;
model: AVANT 4000). Further study will be carried out to
quantify the accuracy of the heart rate measurement using
the FBG sensor.
Figure 12 gives some screen snapshots of respiratory
rate and pressure distribution contour when a person is
lying on the smart bed.
With this real-time pressure distribution monitoring,
when a patient stays unmoved for a long time, it will
automatically alert the nurse to help move the patient’s
body to prevent bedsore generation. In more advanced
behavior tracking, the system is able to detect the position
of the bedridden patient, whether the patient has fallen off
the bed or even to detect the agitation level of a patient.