IEEE TRANSACTIONS ON ULTRASONICS, FERROELECTRICS, AND FREQUENCY CONTROL, . 61, . 5, MAY 2014 792
0885–3010
©
2014 IEEE
A Volumetric CMUT-Based Ultrasound
Imaging System Simulator With Integrated
Reception and μ-Beamforming
Electronics Models
Giulia Matrone, Member, IEEE, Alessandro S. Savoia, Member, IEEE, Marco Terenzi,
Giosuè Caliano, Senior Member, IEEE, Fabio Quaglia, and Giovanni Magenes, Member, IEEE
Abstract—In modern ultrasound imaging devices, two-di-
mensional probes and electronic scanning allow volumetric im-
aging of anatomical structures. When dealing with the design
of such complex 3-D ultrasound (US) systems, as the number
of transducers and channels dramatically increases, new chal-
lenges concerning the integration of electronics and the imple-
mentation of smart micro-beamforming strategies arise. Hence,
the possibility to predict the behavior of the whole system is
mandatory.
In this paper, we propose and describe an advanced sim-
ulation tool for ultrasound system modeling and simulation,
which conjugates the US propagation and scattering, signal
transduction, electronic signal conditioning, and beamforming
in a single environment. In particular, we present the architec-
ture and model of an existing 16-channel integrated receiver,
which includes an amplification and micro-beamforming stage,
and validate it by comparison with circuit simulations. The de-
veloped model is then used in conjunction with the transducer
and US field models to perform a system simulation, aimed
at estimating the performance of an example 3-D US imag-
ing system that uses a capacitive micromachined ultrasonic
transducer (CMUT) 2-D phased-array coupled to the modeled
reception front-end. Results of point spread function (PSF)
calculations, as well as synthetic imaging of a virtual phantom,
show that this tool is actually able to model the complete
US image reconstruction process, and that it could be used
to quickly provide valuable system-level feedback for an opti-
mized tuning of electronic design parameters.
I. I
A
ultrasound (US)-based imaging systems at
the state of the art allow 3-D real-time (4-D) visu-
alization of moving anatomical structures [1], [2], thanks
to two-dimensional phased arrays of piezoelectric trans-
ducers and electronic volume scanning [3], [4]. In these
systems, beam generation, focusing, and steering in both
azimuth and elevation directions is entirely managed by
the device electronics [5]; moreover, digital beamforming
is implemented on dedicated very large scale integrated
(VLSI) circuits [6]. This makes it possible to achieve high
frame rates together with dynamic focusing and thus bet-
ter image resolution and depth of field.
The design of a volumetric imaging system still poses
some challenges and critical technological aspects to be
addressed, such as transducer fabrication, arrangement,
and interconnection, the need for a high dynamic range
and very-low-noise front-end reception circuitry, electron-
ics size, power constraints, and higher frame rates for
4-D imaging (e.g., cardiac) applications. To enable 3-D
real-time imaging and to achieve good performance from
an acoustic point of view, in fact, the transducer array
should be spatially sampled at less than one-half wave-
length pitch (a fully populated array). Thus, the number
of transducers and channels must be increased to the or-
der of thousands, even for small apertures, which leads to
practical wiring problems. Connecting all channels to the
main system through coaxial cables, as traditional systems
do, would definitely be unfeasible [7]. Moreover, the direct
coupling between the transducer array elements and the
cables severely reduces the SNR of the received signals.
For these reasons, in modern US scanners, the trend
toward miniaturization is leading to the integration of the
electronics and some signal processing abilities directly
within the probe handle and in close contact with the
transducer array to enable 3-D imaging [5], [8]. This ten-
dency can be further facilitated by MEMS-based capaci-
tive micromachined ultrasonic transducer (CMUT) tech-
nology [9]. CMUTs are very appealing because of their
high compatibility with microelectronic technologies [10],
which makes their employment (in certain applications)
preferable to traditional piezoelectric technology [11]–[13].
A smart solution for addressing this challenge and par-
tially solving wiring problems is to reduce the effective
number of channels to be routed to the main imaging
system, by implementing and integrating a first stage
of beam-formation [called micro (μ)-beamforming] on a
front-end application-specific integrated circuit (ASIC)
placed inside the probe, close to the transducer array [14]–
[17]. Basically, the μ-beamformer is in charge of delaying
and summing the signals received by the transducers, af-
ter being properly conditioned. μ-beamformers operate a
fine realignment of the echoes received by each small sub-
aperture inside the array (Fig. 1); then, at a higher level,
the main beamformer combines these signals into a unique
Manuscript received December 12, 2013; accepted February 19, 2014.
G. Matrone and G. Magenes are with the Dipartimento di Ingegneria
Industriale e dell’Informazione, Università degli Studi di Pavia, Pavia,
Italy (e-mail: giulia.matrone@unipv.it).
A. S. Savoia and G. Caliano are with the Dipartimento di Ingegneria,
Università degli Studi Roma Tre, Rome, Italy.
M. Terenzi and F. Quaglia are with STMicroelectronics, Cornaredo
(Milan), Italy.
DOI http://dx.doi.org/10.1109/TUFFC.2014.2971