IEEE Robotics & Automation Magazine - September 2020 - 52

Results and Discussions
Force Capabilities of Robotic Instruments
As shown in Figure 8, the grasper has a much higher gripping force than the needle driver, because the former is
shorter and has fewer TSMs inside. From C1 to C6, the
output force decreases, because the mechanical advantage
decreases, and the joints are in increasingly critical angles
in these configurations. However, the gripping forces of
the needle driver and grasper are about 2.5 and 4.8 N even
at the most critical configuration (C3), and they can be 4.3
and 5.8 N when the arms are straight (C1). These capacities are sufficient for most cases, as discussed in the following section.
6
Needle Driver
Grasper

Force (N)

4
2
0

C1

C2

C3

C4

C5

C6

Figure 8. The output force capacities of the robotic needle
driver and the grasper at different configurations. C1~C3 are for
gripping forces, C4~C5 are for yaw steering forces, and C6 is
for pitch steering forces (see the section "Force Capabilities of
Robotic Instruments" for detailed definitions of C1~C6).

Colon

3

Puncture
Puncture
Point
Punctu
cture
ctu
re Poi
int
nt
2 Pun
Point
1
0

Stomach

4
Force (N)

Force (N)

4

3
2
1
0

0
2
4
Displacement (mm)

0
2
4
Displacement (mm)

t = 1 mm
t = 2 mm
t = 3 mm

t = 2 mm
t = 3 mm
t = 4 mm

1
0.5
0

(b)
Force (N)

Force (N)

(a)
1.5

1 2 3
Colon Thickness (mm)
(c)

4
2
0

2 3 4
Stomach Thickness (mm)
(d)

Figure 9. Puncture tests on ex vivo pig colon and stomach
tissues. The force trajectory examples (with respect to tendon
displacements) for (a) colon tissues and (b) stomach tissues.
Averages and standard deviations of the puncture forces of (c)
colon tissues and (d) stomach tissues.

52

*

IEEE ROBOTICS & AUTOMATION MAGAZINE

*

SEPTEMBER 2020

Puncture Force of Ex Vivo Tissues
As shown in Figure 9(a) and (b), when the tendon displacement increases, the upper jaw closes, and the force applied by
the needle to the tissue increases; once the tissue is fully penetrated (puncture point), the force suddenly drops. Then, the
force rises again as the needle tip reaches the mechanical limit.
As the tissue gets thicker, the average puncture force increases
[Figure 9(c) and (d)]. The puncture force for stomach tissue is
normally higher than that for colon tissue since the former is
generally stronger (or tougher) than the latter. The measured
puncture forces for colon and stomach tissues are in the range
of 0.6-3.4 N, which is within the force capability of the needle
driver. These are only ex vivo measurements on pig tissues. More
accurate data need to be collected in vivo (ideally, in human bodies) with small force sensors [12], [13] on the needle driver.
The sudden drop of the force at the puncture point
always comes with a snapping motion of the upper jaw,
which the user can also see through the endoscope camera
during suturing. The snapping motion means that the tissue
is punctured, and the needle quickly reaches the mechanical
end and thus is ready to be switched to the opposite jaw.
This visual clue is critical for successful needle swapping and
has been used during the following in vivo study.
In Vivo Animal Trials
The suturing process is shown in Figure 10. The needle
driver on the right can accurately point the needle tip to
the desired stitching point [Figure 10(b)] and to the
suture loop [Figure 10(d)]. Meanwhile, the grasper on the
left played an important role in lifting and feeding tissue
to the needle driver as well as handling the suture thread.
The time for making the stitches and creating the knot
was 11 and 4 min, respectively. For details, refer to the
supplementary video found in IEEE Xplore.
The needle-locking mechanism was reliable and ensured
successful needle switching during the trial. When the needle
penetrated through the tissue, the rotating jaw always had a
snapping motion (see the section "Puncture Force of Ex Vivo
Tissues"), and the needle tip became visible, both of which
indicated that the needle was ready to be switched. Apart from
the pitch and yaw DoF, we also found that the rolling DoF was
highly important as it helps the end effectors approach the tissue from a proper angle, which is necessary for incisions with
arbitrary orientations. The through-the-scope feature is particularly useful when a new needle needs to be used for additional stitches. The feasibility of deploying the needle using a
nitinol guidewire was also confirmed to be robust. Additionally, we confirmed that, as illustrated in Figure 3, puncturing tissue with the needle on the lower jaw is much more efficient
than puncturing with the needle on the upper jaw.
With the robotic suturing system, suturing is performed
by two cooperative robotic instruments, a technique that
naturally suits the suturing tasks. This made suturing with
this system easier and more intuitive compared with other
devices. In addition, this system supports endoscopic
knot tying (not possible with other devices) and is also
IEEE Robotics & Automation Magazine - September 2020

Table of Contents for the Digital Edition of IEEE Robotics & Automation Magazine - September 2020
