IEEE Robotics & Automation Magazine - September 2023 - 83

immersed in water. Additionally, flexibility and dexterity
were limited due to the reduced number of DoF. A 5-DoF
parallel link robotic HIFU system was proposed by Tang et al.
[14]. The system was designed for breast treatments, and it
enabled 3D US image reconstruction, target segmentation,
treatment path generation, and automatic HIFU irradiation.
No coupling device was presented, and an underwater workspace
was designed for therapy delivery, thus potentially limiting
patient comfort and system flexibility. A robot-assisted
HIFU system for the treatment of peripheral artery disease
was reported in [15], yet the robotic arm was simply used as
a static holder.
However, in all the previous studies, no tracking strategy
to enable the treatment of moving organs was implemented.
Koizumi et al., in [16], reported a USgHIFU system that compensated
for motion by tracking and following the area to be
treated by stereo US imaging. The transducer was completely
immersed in water, and it was moved by a motorized slide,
thus limiting the flexibility of therapy delivery. Chanel et al.,
in [17], described the only robotic USgHIFU system for realtime
compensation for intra-abdominal organ motion by US
visual servoing. However, this strategy was able to compensate
for only 82.6% of the breathing-induced motion, and it
neglected the patient-transducer coupling during motion
compensation. An et al. [18] introduced a US imaging-assisted
robotic HIFU ablation system with a displacement compensatory
mechanism. Target motion compensation was based on
a correlative polynomial function between the heaving chest
position, recorded using an optical tracker, and the tumor
position, recorded using US imaging. However, this strategy
showed a compensation error of 1.72 ± 1.26 mm. In addition,
the robotic arm was a 4-DoF manipulator, and the system was
designed to be fully immersed in water, thus reducing patient
comfort and system dexterity and flexibility.
OBJECTIVE OF THIS WORK
This literature overview highlights the lack of a robot-assisted
HIFU platform that takes real advantage of robotics to
cover all the limitations of the currently available HIFU systems.
In this direction, some years ago, we introduced the
Focused Ultrasound Therapy Using Robotic Approaches
(FUTURA) platform [19]. This platform combines two independent
robotic manipulators, one for therapy delivery and one
for therapy guidance. Starting from that experience and with
the aim to improve platform usability, cost-effectiveness, and
portability, we introduce the HIFUSK platform. HIFUSK
combines the HIFU technology developed in FUTURA with
the KUKA LBR medical robot (LBR Med 14 R820), the only
robotic arm available on the market that is certified for clinical
applications. The aim of this work is to develop and extensively
validate a HIFU platform with robotic control and US imaging
guidance to address both static and moving organs through a
dedicated motion tracking technique. A specific focus on strategies
to ensure patient safety throughout the treatment is also
key: in this regard, dedicated components of the HIFUSK platform
are presented and discussed in this work.
DESIGN AND IMPLEMENTATION OF HIFUSK
This section first illustrates the HIFUSK platform in terms of
its components. Then, the description focuses on the system's
key functionalities as a robot-assisted surgical platform for
FUS treatment of both static and moving organs in the
abdominal area.
HARDWARE MODULES
The HIFUSK components are presented in Figure 4(a).
HIFUSK is based on a 7-DoF robotic arm. In all its seven
axes, the LBR Med is equipped with joint torque sensors that
enable the detection of applied forces. The robotic arm
mounts the therapeutic HIFU transducer. This is a 16-channel
phased annular array transducer with a central frequency
of 1.2 MHz, diameter of 120 mm, and focal length of
120 mm (produced by Imasonic). The focal length can be
further adjusted by applying electronic axial steering in the
range ±20 mm, thus obtaining a focal length in the interval of
100-140 mm. The HIFU transducer is driven by a 16-channel
high-power signal generator (Image Guided Therapy).
Each channel can reach up to 20 W, thus generating a maximum
power of 320 W.
To guide the therapy, a 2D linear phased-array US probe,
with a 4-7-MHz bandwidth (Analogic Ultrasound), is confocally
mounted on the HIFU transducer. A research US system
is connected to the US probe for common B-mode US
image acquisition, system parameter selection, and recording
of raw radio frequency data. This final operation is essential
for further imaging analysis and the implementation of
image-based algorithms.
The acoustic coupling between the HIFU transducer and
patient body is ensured by a balloon made with a flexible latex
membrane (150 mn in thickness). The balloon is attached to
the HIFU transducer and filled with degassed water. Water is
continuously recirculated through a degassing circuit to avoid
cavitation effects due to the possible presence of air particles
inside the water-filled balloon while preventing excessive
heating of the balloon itself.
Finally, HIFUSK is equipped with dedicated software (see
the " Software Architecture " section) running on an Intel Core
i7-9700, 3-GHz computer with a NVIDIA GeForce RTX 2070
SUPER video card for synchronization and data exchange
among the different platform components. A touchscreen
monitor (FLATRON 19MB15T-B; LG Electronics) is used to
provide the operator with a GUI.
SOFTWARE MODULES
This section illustrates the modules of the HIFUSK platform
as functionalities that enable robot-assisted FUS treatment.
These functionalities, together with their methodological
details, are described in the order in which they would be
used by the physician throughout the treatment. A video
showing a demonstration of the HIFUSK platform can be
found as a multimedia attachment (available at https://doi.
org/10.1109/MRA.2022.3188221). The safety measures
implemented in the HIFUSK platform will also be discussed.
SEPTEMBER 2023 IEEE ROBOTICS & AUTOMATION MAGAZINE
83
https://doi.org/10.1109/MRA.2022.3188221 https://doi.org/10.1109/MRA.2022.3188221
IEEE Robotics & Automation Magazine - September 2023

Table of Contents for the Digital Edition of IEEE Robotics & Automation Magazine - September 2023
