Biomedical Engineering Reference
In-Depth Information
analysis [7] shows that for
E
L
/
E
T
>
10, the variation of this ratio has a negligible effect on
the output. On the other hand, for the absolute majority of tumors, the reported stiffness
is greater than this ratio [6]. Therefore, in the practical range, the output response is not
much influenced by variation in the Young's modulus of the lump. Another influential
factor is the magnitude of the applied load. The contact force between the grasper and
the tissue depends on the load exerted by the grasper jaws on the tissue. Therefore, it
is necessary to measure the total applied load as well as the pressure distribution. The
applied load can be measured in different ways. For instance, a strain gauge attached to
the jaw can provide the data on the magnitude of the applied load. Another approach to
measure the applied load was presented in Chapter 4 in which an extra PVDF film is
used at the supports of each sensing element [8]. In the experiments conducted in this
study, the load was measured using a reference load cell. Furthermore, to reduce the
number of contributing parameters, we considered force as being a constant factor. The
other remaining factors are the size of tumor and its location in the
x
and
y
directions.
Since the majority of masses can be approximated as spherical features, the number of
parameters to characterize the size of the sensor can be reduced to one value, that is, the
lump radius. The first design (see Figure 8.14) overlooks the depth of lump and locates
a lump merely in the
x
direction (the grasper length). However, by using two sets of
arrays of sensing elements in the second design described in next section, it is possible
to determine the depth of the lump as well.
8.3.2.1 Graphical Representation of Localized Lumps in One Dimension
As shown in Figure 8.14, in the first design the lower jaw is equipped with a sen-
sor array, hence the upper jaw only applies compressive load to the object containing
lumps. To graphically represent the location of the lump, an image with seven vertical
parallel bands corresponding to the seven sensing elements was initially considered (see
Figure 8.16b). The intensity of each band was considered to be proportional to the output
of the corresponding sensing element. The voltage distribution along the sensor array can
be considered as a vector
{
V
}
1
×
7
that is related to the intensity vector
{
I
}
1
×
7
by:
V
α
I
i
=
.(K
−
1
)
|
V
i
≤
α
,
i
=
1
, ... ,
7
(8.8)
I
i
=
K
−
1
|
V
i
>α
where
α
is the normalizing factor that determines the softness sensitivity (Very Soft, Soft,
Medium, etc.), and
K
is the number of grayscales that are used to construct the graphical
image (here
K
= 256). It can be seen from Equation 8.1 that for a given
α
,when
V
i
≤
α
,
the scaling factor
α
maps the input voltage domain onto interval [0, 1], then this value,
using a factor of (
K
- 1), would be mapped onto the corresponding grayscale level,
between 0 and 255. Once
V
i
>α
, all the values of
V
i
would be mapped to the maximum
intensity (i.e.,
I
i
= 255). For example, Figure 8.16b shows the graphical display for a
case where two lumps were detected in the grasped tissue. In this case, one of the lumps
had been positioned above sensing element 6 and the other had been placed above and
between sensing elements 2 and 3 (see Figure 8.16a for the configuration). However, due
to the limited number of sensing elements, the quality of the image shown in Figure 8.16b
was not satisfactory. Therefore, by using an interpolation technique, the quality of the
