Intelligent Signal Analysis Using Case-Based Reasoning for Decision Support in Stress Management - Computational Intelligence in Healthcare: Advanced Methodologies - page 171

For the old case, the input -10.9 can be represented with an mf grade of 1 and the
lower and upper bounds -16.35 and -5.45 with an mf grade of 0.
Fuzzy intersection is applied between the two fuzzy sets to get a new fuzzy set
which represents the overlapping area between them. Similarity between the old case
( S f ) and the new case ( C f ) is now calculated using Equation 2:

(2)

sim ( C , S ) = s ( m 1 m 2 = max( om / m 1 om / m 2

f f f

where m1 , m2 and om are the area of the corresponding fuzzy sets. For instance,
m 1 = 5.45 and m 2 =3.15 where area is defined by the equation area=(1/2) x base x
height. For om= 0.92, height is defined from the intersection point of the two fuzzy
membership functions. So from Equation 2, the calculated local similarity can be min
(0.17, 0.29)=0.17 and max is 0.29. If the mf s are considered as 100 % fuzzified then
minimum local similarity will be 0.34 and maximum will be 0.58. Thus, a user can
get the options both for tuning the mf s and choosing the min/max values for the simi-
larity function based on their requirements. When the overlapping areas become big-
ger, then the similarity between the two features will also increase, and for completely
matched features similarity will be 1.
The system can provide matching outcomes in a sorted list of best matching cases
according to their similarity values in three circumstances: when a new problem case
is matched with all the solved cases in a case base (between subject and class), within
a class where the class information is provided by the user and also within a subject.
Thus, the CBR system provides the classified FT measurement as an output for the
diagnosis of individual sensitivity to stress.

5.1 Calculating the Degree of Changes and Extracting the Features

Together with clinicians we have agreed on a standardization of the slope to make
changes visible and patients and situations easier to compare. The proposal is that the
X axis displays minutes and the Y axis degrees Celsius, hence a change during 1
minute of 1 degree gives a “degree of change” of 45°. A low angle value, e.g. zero or
close to zero indicates no change or a stable finger temperature.
A high positive angle value indicates rising finger temperature, while a negative
angle, e.g. -20° indicates falling finger temperature. Usually, the purpose of step 1
(the baseline, shown in Table 3) is to stabilize the finger temperature before starting

+10° +20° +40°

stable

Steep Increase/
decrease

Too steep, if decrease,
stop test

Increasing/ decreasing
fast

±0°

-10°

-40°

-20°

Fig. 21. Example of visualizations of temperature change, X-axis minutes, Y-axis in degree
Celsius