Page 128 - Big Data Analytics for Intelligent Healthcare Management

5.8 RESULTS, INTERPRETATION AND DISCUSSION 121

EMGv: EMGa showed results in a similar spirit but the improvement rate was continuous in nature
from start to end. This technique/therapy reduced the overall TTH duration per cycle from 90.3 to
76.5min per cycle per subject, i.e., overall improvement of TTH duration per cycle was 13.8min
per subject.
From the above analysis, we can conclude that GSRv showed an overall better improvement in
reduction of TTH duration per cycle per subject when compared to EMGv. But the rate of improvement
in TTH duration per cycle using GSRv in the last 6months was nearly stagnant, which diminished the
probability of further reducing the TTH duration per cycle. On the other hand, EMGa showed a sig-
nificant rate in the last 6months, giving the possibility of decreasing it further. Therefore, in the longer
run, EMGv will be more beneficial therapy than GSRv.
[31] The biofeedback for GSR can be detected by means of a GSR2 biofeedback device. It works by
measuring the moisture of the hand and immediately gives responses in the form of tone. Hence, the
data that comes through the sensors of EMG and GSR is in the form of big data. Since each and every
sensing device is now connected to the Internet, in the coming future, the data that is stored in the form
of big data analytics can be used as a biofeedback in the stress and Tension Type Headache (TTH). The
scope of IoT and big data will increase as the data processing methods have to be replaced by some
other upcoming techniques.

5.8.11 TREND ON CORRELATION OF TTH DURATION AND INTENSITY
This analysis was performed to establish the relationship of duration of TTH pain of the subject with the
intensity of pain. Chronic intensity pains generally stay for longer durations. Fig. 5.18 reveals the in-
teresting analysis of the subjects’ TTH pain duration in correlation with TTH pain intensity. The trends
observed are as follows:
On analysis, it was found that the baseline data for both the groups was mostly of high TTH intensity
and high TTH duration and the correlation between these parameter was observed as high. After ap-
plication of a variety of trend models such as linear, logarithmic, exponential, polynomial, and power
model, we found the best fitted trend in power models. Hence the power model trend has been analyzed.
The mathematical modeling of the power model is given as: (Tables 5.8–5.10).
Analysis of the trend found with the respective techniques are as:
GSRv: The improvement trend of duration of TTH pain with the pain intensity was exponential
almost throughout the period, having an exception at the end after using the GSRv as feedback therapy.
At the end of the period, the relationship became linear, which shows that further improvement in in-
tensity will linearly reduce during the duration period.
Also, the analysis showed that the majority of the data came under the quadrant of low duration and
low intensity with the exception of four subjects out of which three were having low intensity but high
duration headaches and one was having low duration but high intensity headaches at the end of 1year.
EMGv: There were dramatic results found in the trend just after applying this technique for
1 month. The trend became linear after this period and remained linear thereafter. This resulted in
two subjects falling outside the quadrant of average low duration and average low intensity. These
two subjects were lying in the low intensity but high duration quadrant after the end of the year.
We can conclude that EMGv has a better improvement correlation between duration and intensity.