Keywords
ultrasound - elastography - history
Introduction
Where are we coming from and where are we going [1]
[2]? Developments in ultrasound
imaging as we know them today are not self-evident but have undergone a continued
and sustained evolution. WFUMB has a website showing the history of ultrasound
worldwide [https://wfumb.info/pdfhistory-2/]. In addition, EFSUMB has prepared an
online e-book on the history of ultrasound, which displays the history of its
members [https://efsumb.org/history/]
and is a work in progress. The “History of ultrasound in medicine from its birth to
date on the occasion of the 50-year anniversary of EFSUMB” [3] and the “History of ultrasound in obstetrics
and gynecology from 1971 to 2021 on the occasion of the 50-year anniversary of
EFSUMB” [4] have been published. An editorial
and booklet also summarize the birth and 50 years of EFSUMB [5]
[6]. In
addition, the history of contrast-enhanced ultrasound has been summarized [7]. Highlights of ultrasound over the past 70
years were summarized by Nielsen et al. [8].
The history of student ultrasound education (SUSE) is also being summarized in an
ongoing study. The current article focuses on the history and development of
elastography in ultrasound and how it became a major imaging modality in medical
diagnosis. What followed the more than 2000 years of manual palpation?
Manual palpation during physical examination has been used in clinical practice for
many centuries to distinguish harder areas from softer ones in the skin, organs,
and/or tissues and to detect tumors. Palpation reflects various characteristics of a
certain structure including the position but also other features, including
hardness, mobility, and pulsation. Palpation is subjective and mainly restricted to
superficial anatomical structures or larger and deeper organs and neoplasia. In the
Western World, the earliest description of this method was found in the Edwin Smith
Papyrus scroll containing an ancient Egyptian medical text. It is believed that the
author of this text was Imhotep, a priest, physician and presumed architect who
lived between 3000 and 2500 B.C. A description of palpation techniques and results
is also found in the Corpus Hippocraticum (Greece), a collection of ancient medical
texts dating from the 5th to the 2nd century. Similar texts are found in the
literature of the Far East but these have not been attributed to a single
author.
Manual palpation still plays an important role in clinical practice, particularly in
special situations, e. g., rectal digital examination of the prostate gland,
intraoperative detection of certain structures, in self-examination of the female
breast. However, it is a rather subjective method and is highly dependent on the
experience of the person performing it. Elastographic methods are not a mere
substitute for palpation but an additional technique that may allow further tissue
characterization. The availability of semi-quantitative and quantitative methods to
determine tissue stiffness allowed a real breakthrough in the diagnostic workup of
patients including freehand, step and physiological displacement, low-frequency
vibration, and radiation force techniques including acoustic emission shear wave and
other impulse and displacement techniques.
Early history of elastometry and elastography
Early history of elastometry and elastography
Conventional ultrasound techniques are limited to image anatomy, the evaluation of
tissue motion in real time, and the display of vessel anatomy blood flow mainly by
Doppler techniques. In addition, emerging techniques exploit elastic properties of
tissue including stiffness or hardness. The first elastometric and elastographic
measurements using Young’s or shear modulus were performed following classic
mechanical models with the application of pressure and tension. From the physical
point of view, elastography quantitatively maps Young’s modulus, which corresponds
to stiffness and shows the mechanical stress and strain of the material using a
simplified approach via Hook’s law [9].
However, as mechanical stress is not easily measurable in vivo, a strong assumption
of homogeneous mechanical stress was initially assumed. Yet, considering different
tissue structures within the human body, the application of pressure and tension as
a homogeneous stress factor is not easily attainable [10]
[11]
[12]. In the beginning, simple
(linear) properties of the underlying material were assumed with viscous and
non-linear properties not taken into account.
Elasticity in general is described as the tendency of whatever material to resume its
original size and shape after deformation by any kind of force or stress. Change in
size or shape are defined as strain, which might be expressed as a ratio of a
diameter (e. g., length) or as a histogram. Solid structures resist changes in shape
and volume since they possess rigidity or shear elasticity, as well as volume
elasticity. In contrast, fluids resist a change in volume, but not in shape as they
possess only volume elasticity. The measurement of elastic properties (elastometry)
of biological tissues was first described by King and Lawton. In their experiments,
resonance was created with mechanical vibrations. They postulated that the relative
elasticity of the examined tissue could be obtained [13]. Using tension experiments, the tissue was described as a
mechanically non-linear, anisotropic, and inhomogeneous viscoelastic medium [14]. Mechanical tissue properties were examined
by determining the elasticity, particularly in muscle and heart samples [15]
[16]
[17]
[18]
[19].
In these experiments, a correlation between tissue elasticity and the acoustic
reverberation coefficient was found [20]. The
measurement and interpretation of superimposing waves have also been used for
elastography of non-linear tissue properties [21], such as in femoral muscles using Doppler ultrasonographic techniques
and in a model using a vibrator-based model with a mechanically moving piston [22]. The description of these techniques and
their corresponding biological tissues was characterized by a lack of uniformly
defined terminology.
Magnetic resonance elastography (MRE) was initially developed at the Mayo Clinic,
Rochester, MN, in 1995 [23]
[24]
[25]
[26]
[27]. In this modality, a static shear modulus
is calculated from the phase shift of the MR signal and the corresponding sites in
three dimensions with high-speed volume acquisition. The extended scan times of MRE
and the application of vibrators to generate appropriate shear waves were
problematic due to the high demand regarding operation time [28]
[29]
[30]
[31]
[32]
[33]
[34]. The advantages include that MRE is not
restricted by the presence of air, gas, or bone, whereas the advantages of USE
include that elastometry of soft tissue strain and elasticity is faster, more
convenient, more accurate, and more precise. For an extensive review of the history
of ultrasonic elasticity imaging technology development, refer to the articles by
Wells and Parker [9]
[35].
Definitions, Elastography, and Elastometry
Elastography is a qualitative method and describes the pictorial representation
of tissue stiffness. Elastometry is a quantitative method with objective
measurements. Currently, two main sonographic elastography techniques are used
in medicine: compression elastography (also called strain or static elastography
or simply elastography) and shear wave elastography (elastometry). The term
“static elastography” was used in contrast to dynamic elastography (MRE,
sonoelasticity) and transient elastography. Shear wave elastography is
quantitative, and compression or static elastography is qualitative or
comparative (harder or softer compared to the environment). Compression
elastography is particularly useful for characterizing circumscribed changes,
and shear wave elastography is useful for studying diffuse (infiltrative)
changes of parenchymal organs.
Ophir and colleagues in Houston, Texas, USA first used the term “elastography” in
1991 [36]. This group was the first to
present elastographic images comparing tissue stiffness in following
publications [37]
[38]
[39]
[40]
[41]. External pressure was used to compress
the tissue and the subsequent deformation was recorded using ultrasound with
cross-correlation between the pre- and post- compression signals. Using a simple
mechanical model, the modulus of elasticity can be derived from the measured
strain and the applied stress (Hooke’s law). The tissue strain was estimated
from the deformation which was combined with assumptions about tissue geometry,
homogeneous stress, and material models through image reconstruction algorithms
that generated images of the underlying qualitative material elastic moduli, the
so-called elastogram [36]. As subsequent
elasticity imaging methods were introduced, this method became known as
“compression elastography”. Several years passed before compression elastography
became commercially available. In fact, due to challenges with computation time
and material model assumptions, commercial implementations do not generate
“elastograms”, but rather bypass the reconstruction step and provide images of
the tissue strain. As such, commercial implementation of compression
elastography is broadly called “strain imaging”. Strain imaging became
commercially available in the early 2000’s (i. e., “real-time elastography” by
Hitachi, Tokyo, Japan was introduced in 2003). Siemens also introduced its
‘‘eSie Touch Elastography Imaging’’ with accepted clinical applications focusing
on the differentiation of benign from malignant lesions in the breast and the
prostate. The stiffness of the tissue is shown on a color image superimposed on
the standard B-mode ultrasound image [10].
Since that time, all major ultrasound manufacturers have provided strain imaging
features, and many other clinical applications have been explored and thousands
of studies have been published.
Using a linear tissue model, stiffness parameters were postulated and derived
from the observed tension in the target tissue. Cross-correlation was used to
compare the measurements in the native and compressed states. A major
disadvantage was the linear estimation function of the underlying tissue model,
the assumption of spatially homogeneous stress, and the high number of measuring
procedures. The first trials were performed in human muscles and breast [42].
An anecdote about the color schema used in strain elastography
The attempt by scientific experts to call for a manufacturer-independent
standardization of color coding at the EFSUMB guideline conference on
elastography in Bologna in September 2012 was not able to prevail against
the color associations of the company representatives, which had traditional
meanings. For the Japanese manufacturing company Hitachi, “red” (the red
circle in the center of the Japanese flag, the sun) symbolized vitality and
strength, and thus has positive emotional connotations as the color of life.
Thereby the color “red” stands for the sun and for the qualities clear,
bright, brilliant, vital, benign and thus “soft(er)”. For manufacturers from
Western cultures (for example, Supersonic Imagine), on the other hand, red
is a color associated with aggressiveness and danger and is understood as a
warning signal in a medical and technical context (fire). Therefore,
ultrasound manufacturers from Europe and the USA code the tissue property
“hard” with the color “red” in their elastography modules [43]. The manufacturer-dependent
difference in the colors used to denote “stiffer” or “softer” must be taken
into account during the examination.
Ultrasound elastometry
This definition is focused on measurements. The term was used in the early
years of development but has now been replaced by the term elastography.
Freehand elastography
The development of freehand elastography was achieved by palpating the body
using a hand-held transducer. The very first related ideas and reports were
published by Christopher R. Hill [44],
who inspired Jeff Bamber to explore this line of work in elastometry when he
was a young postdoc [45]
[46]
[47]. The abstracts were presented at the World Federation for
Ultrasound in Medicine and Biology conference Washington DC, United States
(1988). This ended up being the strain imaging method implemented on most
commercial systems, and differed from Ophir’s original concept of strain
imaging, which used a motor-driven transducer. Eventually Ophir stopped
developing that approach and switched to the freehand method introduced by
Bamber et al. [48].
The work done by Bamber et al. then developed in two directions [49].
-
A combined autocorrelation algorithm for overcoming computer speed
limitations, first published by Shiina et al. [50], was subsequently further
developed and commercially released by Hitachi in about 2003.
-
A general strain imaging algorithm for freehand use mentioned in two
prize-winning conference presentations [51] (winner of the BMUS oral
presentation prize) and [52]
(winner of the RSNA INFO-RAD presentation certificate of merit)
eventually became the algorithm commercially released by Zonare
Ltd., in about 2008, on their Z.one Ultra scanner [53].
Sonoelastography
The terms sonoelastography and dynamic elastography broadly refer to dynamic
elasticity imaging methods that use harmonic vibrations (with fixed
frequencies, which differ from transient excitation) to mechanically excite
tissue and reconstruction algorithms to estimate tissue material properties
and ARFI-based techniques. The term sonoelastography was the result of a
collaboration between the NIH (Rehabilitation Division) and the Tokyo
Institute of Technology. Elasticity measurements were performed on patients
and compared [54]. Krouskop et al.
(1987) reported one of the first quantitative measurements of tissue
elasticity using gated pulsed Doppler to track the velocity of internal
tissue displacements generated by an externally applied vibration [22]. These first 1D noninvasive
measurements of the elastic modulus of soft tissue in vivo were aimed at
predicting the interaction of a prosthetic socket with an amputee’s residual
limb.
Vibration amplitude sonoelastography was introduced by Lerner and Parker in
1987–88 [55]
[56]
[57]
[58]
[59]
[60]. Robert Lerner, a radiologist at the University of Rochester,
had the experience of palpating prostate carcinomas that were not visible on
ultrasound [35]. Subsequently, in
1987, Lerner and Parker conducted preliminary studies on vibration amplitude
sonoelastography (sonoelasticity imaging) [57]
[61]. The principle
involved applying a low-frequency vibration (20–1000 Hz) externally to
initiate internal vibrations in the tissue under investigation.
External vibrators generated shear waves in tissues, which were then
monitored using Doppler sonography [57]. Transient elastography was introduced by Mathias Fink and
collaborators from 1997 at the Laboratoire Ondes et Acoustique (now Institut
Langevin) to overcome the influence of boundary conditions observed in
sonoelastography.
Measurement of the stiffness of living tissue using ultrasonic radiation
force
Measurement of the stiffness of living tissue using ultrasonic radiation
force
Sugimoto’s group from the Tokyo Institute of Technology investigated the
possibilities of tissue hardness measurements using the radiation force of focused
ultrasound (ARFI) [62].
The authors used the force of a focused ultrasound beam to assess the longitudinal
displacement of the tissue (like in strain elastography). A major possible
disadvantage is that different elastic waves show significant interactions, such as
conventional longitudinally orientated ultrasound waves (1500 m/s) modified by
compression may be superimposed on the much weaker shear wave characteristics
(1 m/s). However, in this study, the measurement is performed while avoiding the
time zone affected by this shear wave after the radiation force is stopped. There is
almost no effect on the displacement measurement in the focal region [63]
[64]
[65]
[66].
The deformation is measured by the pulsed Doppler method. As an advantage, this
technique makes it possible to evaluate tissue hardness based on only the time
dependence of relative displacement. By using the Voigt model, the viscoelastic
value is derived from the change in deformation. Two-layer agar samples were
employed, and a fundamental experiment was carried out to obtain the viscoelastic
diagrams. These diagrams demonstrate the feasibility of the hardness measurement.
The study was conducted between 1988 and 1993. For the first time, theoretical and
experimental studies of in vivo measurement about tissue hardness were
performed using focused ultrasound.
In 1988, in response to the request to display more qualitative information, such as
hardness information, rather than just grayscale images, Professor Sadayuki Ueha of
Tokyo Institute of Technology gave Tsuneyoshi Sugimoto, who was a master’s student
at the time, “tissue hardness measurement using radiation force” as a research
topic. He continued this research as a doctoral research theme. The first problem
was what was the hardness felt by the medical doctor. However, it was clear that, in
general, biological tissue is a complex and viscoelastic body, and in vivo
measurement is ultimately required. Because of that, no practical measurement method
was established at that time. Therefore, with the cooperation of Professor Kouichi
Itoh of Jichi Medical University, the relationship between the amount of
pressurization and the amount of deformation was investigated for various human
organs. As a result, it became clear that the hardness felt by human beings is
related not only to elastic information but also to viscous information. After that,
T. Sugimoto measured the deformation of tissue in the living body and its temporal
decay due to the radiation pressure of focused ultrasound using the pulse echo
method or the pulse Doppler method. A method to extract the feature quantity related
to hardness from the deformation measurement was proposed. Moreover, in order to
extend the proposed method to in vivo hardness evaluation, relative value
measurement is indispensable. Then, a viscoelastic diagram was devised with elastic
hardness on the vertical axis and viscous hardness on the horizontal axis.
Furthermore, in order to evaluate the viscoelastic value of the biological tissue on
the basis of the time change of displacement, a theoretical study was performed
using the Voigt viscoelastic model, after the biological tissue was deformed by the
radiation pressure of focused ultrasonic waves. As a result of the examination, it
was clarified that the index value corresponding to each viscoelastic value of the
model can be derived from the time change of displacement and the viscoelastic
diagram can be created.
Strain imaging, principles, and basic mechanisms
Strain imaging, principles, and basic mechanisms
During conventional B-mode sonography with grayscale imaging, organs, including
circumscribed lesions within the organs, can be palpated with the transducer by
applying compression. Using this technique, the relative strain (or compression) of
the target lesion can be evaluated in comparison to the surrounding tissues [67]. This can be semi-quantitated using a
defined scale. The aim of this procedure is to gain information regarding tissue
properties such as consistency, rigidity, size, and deformability. Ultrasound
frequencies from 1 to 20 MHz with pulse repetitive frequencies of 1–10 kHz are used.
In strain images, softer tissues will show a higher degree of deformation than
stiffer tissues for a given compression (or applied stress). This deformation is
visualized by ultrasonographic image processing with the examination being performed
using standard ultrasound probes without any additional devices (pressure
measurements, vibrators, etc.) similar to Doppler examination. The calculation of
the strain is performed in real time with the results superimposed onto the
conventional B-mode image [68].
Probe systems, standardization of manual compression
Clinical strain measurement techniques
Relative strain is obtained by measurements and transformed into images with
the extended autocorrelation method by the ultrasound systems. Color coding
with a blending of transparency up to 100% color saturation represents data
translation of the obtained strain values. Using a measuring menu, a region
of interest (ROI) within the elastographic window of the target lesion can
be positioned and relative strain data obtained as the mean (%) of the areas
within the ROI. These values can numerically supplement the color data, as
is the case with thyroid examinations [69]
[70]
[71]
[72]
[73].
Strain Ratio
Fat-to-lesion relationship
Linear correlations between different strain parameters were first
investigated using appropriately sophisticated phantoms with embedded
target structures. These results permit the definition of these
properties more quantitatively in terms of strain ratios. If ROI “A” is
positioned over the lesion to be examined (e. g., a tumor appearing in
blue) and a second ROI “B” is placed over a reference tissue, e. g., fat
deposits, then the lesion A (ROI A) is expected to appear with very low
values (low elasticity) in relation to the reference structure (fat ROI
B) which will show much higher values (high elasticity). In the clinical
environment, these observations correspond to characteristics found in
the human breast. The fatty tissue of this organ can be regarded as the
reference structure from which strain parameters can be determined and
expressed as the ratio B/A (fat-to-lesion), which is a dimensionless
number. If this ratio is around 1.2, a lesion is highly likely to be
benign, whereas for a ratio of greater than 8.5 malignancy is suspected.
This ratio comparing the tissue of interest to fat benefits from the
fact that fat tissue has very similar stiffness values from patient to
patient. The fat-to-lesion ratio reduces other variabilities due to the
experiment and technology. Consequently, the number of false-positive
test results was reduced, particularly for benign lesions. A cut-off of
3.2 was defined [74]
[75]
[76]
[77]
[78]
[79].
Strain Histogram
More recent developments regarding elastography data processing include
signal averaging of the last recorded strain sequence with subsequent
quantitative color coding. With the strain-histology mode, an ROI can be
placed within the elastobox with discrimination of the elastography
data. The difference with respect to the above-described method is that
discrimination of elastographically generated data is possible using
this technique. This ROI is placed over a suspected area (with blue
color elements) and a histogram of the relative stretching in a blue
area is generated within the range 0 (blue) to 255. This ROI is of
statistical importance with regard to size and circumference.
Quantitative values (from 0 to 255) are then available based on the data
derived from appropriate numerical tables. In addition, a statistical
distribution of the non-stretchable areas (blue) can be given. Within
the ROIs blue areas will be recorded in clusters and calculated with
regard to area and circumference, thereby enabling a statistical
description (complexity, standard deviation). This technology is thought
to noninvasively evaluate parenchymatous organs such the liver for
staging fibrosis in addition to the classic gold standard of biopsy.
However, the results that have been obtained in European cohorts do not
allow the technique to be recommended in this setting [80]
[81].
Transient Elastography
Transient elastography began at the laboratory of Mathias Fink, in Paris, France in
the late 1990 s [82]
[83]
[84]
[85]. The principle of transient
elastography is to measure the speed of shear waves generated by a low-frequency
transient vibration generated at the surface of a medium. The displacements induced
in the medium by the propagation of the shear waves are tracked using ultrasound
with a high frame rate (typically higher than 1000 frames per second). Initial
developments by Catheline during his PhD under the supervision of Fink consisted of
a transmission setup in which a mini-shaker and a single element ultrasound
transducer were located on opposite sides of the to-be-measured medium [85]
[86].
It was during the Sandrin PhD that the technique was extended to work in reflection
mode, with the ultrasound probe directly located on the mini-shaker, thus resulting
in a more compact and clinically applicable device. This research was the origin of
the shear elasticity probe [83], while the
diffraction effects associated with shear wave generation were better understood
[87]. The transient elastography technique
was initially a 1D technique. To extend transient elastography to 2D imaging, Fink
had the idea of modifying the electronics of a time-reversal mirror made of an
ultrasonic array (that he developed in the early 90s) to make an ultrafast
ultrasound scanner able to acquire several thousands of images per second to track
the propagation of shear waves in 2D [88].
Fink and Sandrin developed this ultrafast imaging ultrasound scanner [89] to obtain elasticity maps of biological
tissues [67]
[69]. The initial idea of Fink’s group was to put the ultrasonic array on
a low-frequency shaker to both generate the shear wave in the body and to observe it
with the ultrasound array [88]. However, the
probe was heavy and not easy to manipulate.
They later decided to use ultrasonic radiation force to generate the shear waves
using the supersonic mode. Interestingly, Fink, Sandrin, and Manneville also applied
the ultrafast imaging ultrasound scanner to flow imaging including vortex [90]
[91].
The Princeps patent of transient elastography [Princeps Transient Elastography
patent US 6,770,033] which was filed in 1999 covers different embodiments of the
technology. Initially proposed for the food industry (yogurt and camembert)
applications, transient elastography ultimately became established in the medical
field [Table 1].
Table 1 Modern ultrasonographic elastography according to Cui
et al. 2022 [126].
|
Technique
|
Measurement
|
Stimulation
|
Method
|
Indicator
|
Company
|
System
|
|
Strain imaging
|
Strain or displacement
|
Manual
compression
|
Strain elastography
|
Elasticity score
Strain ratio
E/B size ratio
|
Esaote
Hitachi
Aloka
GE
Philips
Toshiba
Ultrasonix
Mindray
Samsung
Siemens
|
ElaXto
Real‑time tissue
elastography
Elastography
ElastoScan
eSieTouch
elasticity imaging
|
|
Shear wave
imaging
|
Shear wave
speed
|
Mechanical vibration
|
Transient
elastography
|
Young’s modulus (kPa)
|
Echosens
|
FibroScan
|
|
|
ARFI
|
p‑SWE
|
Shear wave speed (m/s)
Young’s modulus (kPa)
|
Siemens
Philips
|
VTQ
ElastPQ
|
|
|
|
2D SWE
|
Shear wave speed (m/s)
Young’s modulus (kPa)
|
Siemens
SuperSonic Imagine
|
VTIQ
SWE
|
|
|
|
3D SWE
|
Shear wave speed (m/s)
Young’s modulus (kPa)
|
SuperSonic Imagine
|
SWE
|
VTQ: Virtual Touch quantification; VTIQ: Virtual Touch image quantification;
SWE: ShearWave elastography; ARFI: Acoustic radiation force impulse; p‑SWE:
Point‑SWE.
Transient elastography in the food industry
The results from applications in the food industry were less successful than
initial expectations [92]. Real-time
assessment of the viscoelastic properties of yogurt during manufacturing was
performed in 1998. Although encouraging results were obtained based on yogurt
purchased at the grocery store, about 15 days after manufacturing, tests at the
manufacturing site were very disappointing. As a matter of fact, fresh yogurt
was ultrasound transparent in the first couple of days, thus preventing any
attempt to perform ultrasound elastography. In the case of meat or muscle,
tissue anisotropy leads to major difficulties triggered by the complexity of the
shear wave propagation. Another problem is the attenuation of ultrasound waves
by fat which was a major obstacle to the noninvasive determination of the stage
of maturation of cheese such as camembert (unpublished data).
Transient elastography in the medical field
In June 2001, Laurent Sandrin and colleagues founded the firm Echosens
(www.echosens.com) to further develop the transient elastography technique which
was still immature. They created algorithms to reliably measure shear wave speed
in vivo and designed an elastography device able to control and synchronize the
low-frequency vibration while performing pulse-echo ultrasound acquisitions. The
improved technique named Vibration-Controlled Transient Elastography (VCTE) was
patented in 2002 [Vibration-Controlled Transient Elastography patent US
7,578,789] [Fig. 1].
Fig. 1 First clinical images of shear wave elastography (SWE)
using ultrafast ultrasound imaging and acoustic radiation force
generation for breast cancer diagnosis.
In parallel, they initiated a marketing survey to identify the potential market
applications. During a meeting together with clinicians, the disadvantages
regarding focal lesions such as breast and prostate cancers were discussed. The
argument was that the differentiation of benign from malignant tissue by
elastography might not be entirely reliable since even benign tissue can show
different degrees of stiffness. A new important criterion regarding the
evaluation of liver stiffness for fibrosis assessment was proposed. As stated in
Hippocrates’ (c. 460 – c. 370 BC) aphorisms, liver fibrosis is a well-known
continuous process clearly associated with increased stiffness. Therefore, the
Echosens team decided to focus their efforts on in-vivo liver parenchyma
stiffness evaluation. The clinical usefulness of transient elastography to
evaluate quite homogeneous and continuously increasing fibrosis accumulation
appeared promising.
The technique was initially tested during a pilot study at Institut Mutualiste
Montsouris (Paris, France) in 2001–2002. Stiffness results obtained with VCTE
were compared with the fibrosis stages determined by pathologists from liver
biopsies [93]. In a following multicentric
study, the first clinically relevant results with evaluation of patients with
chronic hepatitis C were published [94].
In December 2003, FibroScan obtained CE marking and was introduced on the market
as the first commercially available noninvasive quantitative elastography
device.
With FibroScan, shear waves are not only measured but can also be visually
demonstrated. The tip of the probe is made of an ultrasound transducer mounted
on the shaft of an electrodynamic actuator. The tip is actuated by transient
excitation generated by the actuator, which induces shear waves in the tissues.
Interestingly, only the longitudinal component of the shear waves is registered
[87] since the displacements are
measured on the ultrasound axis which is also the propagation axis of the shear
waves. The controlled attenuation parameter (CAP) to detect and grade liver
steatosis was introduced in 2010 [95].
Echosens was the first manufacturer to commercially offer stiffness measurement
and ultrasound attenuation measurements in vivo. Since 2001, the company
continues to develop the VCTE technique which is used in all devices named
FibroScan [10].
Acoustic Radiation Force Impulse (ARFI) based shear wave elastography
Acoustic Radiation Force Impulse (ARFI) based shear wave elastography
ARFI was developed by different researchers and commercially introduced by the
Siemens company for the noninvasive measurement of liver stiffness. Sarvazyan et al.
first proposed radiation force to make modulus maps [96]. Nightingale et al. first demonstrated in vivo use of radiation force
of focused ultrasound to generate images [97]
by measuring, point by point, the local displacement induced in tissue by the
radiation force of a focused ultrasonic beam. Fink’s group then decided to extend
the radiation force concept to directly create a transient shear wave of
high-amplitude to replace the one created by the mechanical shaker in 2D transient
elastography. The two fundamental ideas behind the SSI technique are as follows:
first, they proposed interleaved focused beam transmissions for shear wave
generation and ultrafast plane wave transmissions for simultaneous ultrafast
imaging. Second, they proposed to electronically control the radiation force pattern
to create a fast moving “seismic” source of shear waves. If the source moves at a
supersonic speed, it generates a high amplitude plane shear wave like in the “sonic
boom” created by a supersonic plane. It was a very efficient way to produce shear
waves in tissue. Tanter and Fink patented this SSI technique in September 2002 that
led to the creation of Supersonic Imagine company in 2005. Bercoff, Tanter, and Fink
used ultrafast imaging to detect the shear wave propagation produced by this
supersonic seismic source [98]. This was the
origin of the so-called supersonic shear wave scanner [98]. The supersonic shear imaging technique
utilizes the underlying ultrafast ultrasound imaging technique developed for
transient 2D elastography, but it replaces the vibrator with acoustic radiation
pressure. First clinical validation of shear wave elastography using the SSI
technique was published in 2008 [95] for
breast cancer diagnosis. For a detailed review of the concept of ultrafast
ultrasound imaging and applications in elastography and beyond, refer to [99].
Siemens introduced the first commercial SWE system in 2008. The Aixplorer (Supersonic
Imagine, Aix-en-Provence, France) was then introduced in 2009 by Supersonic Imagine
first for breast applications [98]
[100]
[101]
[102] and other manufacturers
followed [103].
ARFI-based shear wave elastography is an ultrasound technique, which is capable of
measuring tissue elasticity quantitatively and also qualitatively (color-coded) when
using 2D-SWE technology. Frequency-defined push pulses, with higher intensity and
longer duration compared to that used for conventional B-mode images, lead to tissue
vibrations at a region of interest (ROI) contributing to the generation of shear
waves (the higher the frequency, the higher the registered velocity). The velocity
of these shear waves can be determined using the same ultrasound transducer. Based
on certain assumptions, this velocity can be converted to tissue stiffness (kPa)
using the Young’s modulus E=3ρVs2
[9]. As ARFI-based shear wave elastography did not rely on ultrafast frame
rates, the shear wave propagation was not tracked in real time but rather
reconstructed in a stroboscopic manner using several successive radiation force
pushes and different imaging lines. This repetition of radiation force transmissions
resulted in a higher acoustic energy deposit and thus limited the elastography frame
rate and capabilities of ARFI-based shear wave elastography for real-time mapping of
elasticity. McAleavey et al. developed a method called spatially modulated
ultrasound radiation force (SMURF) [104]. In
this method, instead of a sharp focal area, a spatially modulated pattern is used to
create a unique spatial frequency within a tissue region.
Fibrosis staging
It can be summarized that although VCTE has by far the highest level of evidence
(more than 4,000 peer-reviewed publications in 2023), the number of publications
on SWE has increased significantly in recent years due to the rapid evolution
and spread of radiation force-based technologies. Several studies have
demonstrated that both pSWE and 2D SWE techniques have similar accuracy to VCTE
for the evaluation of liver fibrosis [5]
[6]
[103]
[105]. However, VCTE and FibroScan are still considered to be the
reference technique and the reference device, respectively, given the simplicity
of use and high level of standardization of the technology.
2D Shear Wave Elastography
Two-dimensional shear wave elastography (2D-SWE) began in the laboratory of
Mathias Fink, in Paris, France in the late 1990s [82]
[83]
[84]
[85]
[89]. Standard ultrasound imaging was not able to efficiently track
shear wave propagation in 2D due to its limited maximum frame rate of several
hundred images per second. Fink and Sandrin [89] developed ultrafast ultrasound imaging which was initially
obtained by only focusing in receive mode. Frame rates of several thousands of
images per second were obtained and used to track shear wave propagation in 2D
(both longitudinal and transverse tissue displacements) [82]. The absence of transmission focus
yields a reduced penetration depth and lower contrast on B-mode ultrafast
images. To solve this problem, Fink’s group later introduced (in 2009) a
multiple plane waves approach (today called coherent compound plane wave
imaging) to synthesize virtual transmission focusing [106]. Bercoff, Tanter, and Fink further
developed the technology in combination with radiation force for the generation
of shear waves [107]. They first focused
on studies with human breast, liver, and prostate, aiming at differentiating
benign from malignant lesions and later on musculoskeletal applications [108]
[109] and cardiovascular pathologies [110]
[111]. With Jacques Souquet
and Claude Cohen-Bacrie, they founded Supersonic Imagine with other colleagues
in 2005. Both Echosens and Supersonic Imagine originated from Fink’s
laboratory.
Further Developments
Echosens pioneered the field of liver stiffness and ultrasound attenuation
assessments. Other manufacturers are developing or have already developed
software for quantifying US beam attenuation. Preliminary results show that
proprietary technologies implemented in ultrasound imaging systems seem more
accurate than CAP for grading liver steatosis [112]
[113]. Another available
method for quantifying liver steatosis is based on the computation of the speed
of sound. Initial results appear promising [114]
[115]
[116]. The backscatter coefficient also
seems accurate for quantifying liver fat [117]
[118].
Modern elastography in clinical practice
Modern elastography in clinical practice
Currently, the terms “elastography” and “elasticity imaging” are used broadly to
describe imaging methods that provide information about tissue stiffness, including
strain imaging, compression elastography, sonoelastography, acoustic radiation force
impulse imaging, ultrasound shear wave elastography, magnetic resonance
elastography, optical coherence elastography, et cetera.
Presently the following ultrasonic elastographic methods are widely being used in
clinical medicine: strain imaging, ARFI imaging (both provide relative images of
tissue stiffness), shear wave elastography (including vibration-controlled transient
elastography), and cardiac strain imaging [119]. Shear wave elastography is quantitative, whereas strain and ARFI
imaging yield qualitative images (harder vs. softer compared to the surrounding
structures). The qualitative methods have mainly been used to characterize
circumscribed lesions and heterogeneous tissues, whereas shear wave elastography
allows both the evaluation of diffuse infiltrations of organs and circumscribed
lesions [68]
[120]
[121].
Elastography has now found wide application including in pediatric patients with a
large number of publications [122]. In
addition to the assessment of liver fibrosis, the status assessment of breast
lesions and thyroid nodules, multiparametric imaging of the prostate in the search
for prostate carcinoma, there is a variety of other applications including the
status assessment of lymph nodes, the assessment of the strength of any focal
lesions compared to the surrounding parenchyma, the assessment of the age of
vascular thrombi, assessment of the vascular wall, intestinal wall assessment in
Crohn’s disease, and last but not least for neurosurgery of brain tumors [35]
[70].
Shear wave elastography was also recently identified as “the new wave of
cardiovascular biomechanics” [123] since
myocardial stiffness is assumed to be a very relevant biomarker of cardiac
pathologies [124]. Recent studies suggest that
diastolic myocardial elasticity is higher in heart failure with preserved ejection
fraction (HFpEF) patients. Systolic myocardial elasticity was also shown to
noninvasively measure cardiac myocardial contractility. Another challenging
application is the combination of strain elastography and semiquantitative methods
with endoscopic ultrasound.
By quantifying the elasticity of muscles, joints, ligaments, nerves, and soft-tissue
musculoskeletal masses, shear wave elastography is also very promising for the
follow-up of various traumatic and pathologic conditions of the musculoskeletal
system [125].
Elastography in endoscopic ultrasound (EUS)
Elastography in endoscopic ultrasound (EUS)
Endoscopic ultrasound strain elastography is an important investigative component in
the evaluation of focal lesions in the pancreas, gastrointestinal tract, and all
structures in the sonographic window of EUS and lymph nodes [127]
[128]
[129]. Elastography in EUS
evolved with the use of electronic EUS probes. Strain elastography in EUS is not
without controversy. In order to be able to evaluate the findings, investigators
must familiarize themselves with the method and study data. Profound background
knowledge is required. The findings must be viewed in the overall context of B-mode,
color Doppler imaging, and contrast-enhanced harmonic EUS imaging data. In EUS,
strain elastography is used to evaluate the hardness of lymph nodes [130]
[131]
[132]
[133], pancreatic lesions [130]
[134]
[135]
[136]
[137]
[138]
[139]
[140]
[141]
[142], and intramural wall structures of the
gastrointestinal tract. The result of strain elastography in EUS is only one piece
of the overall examination and is subject to some influencing factors. Elastography
is possible with both longitudinal and radial probes. The transducer of the EUS
probe is coupled to the wall with slight pressure. The interesting lesion should be
centered. The elastography window, the region of interest, must be sufficiently
large to detect the lesion of interest in comparison to surrounding tissues. By
comparing the elasticity of the lesion with the surrounding tissue, the strain ratio
can be calculated [131]
[141]
[143]
[144]. Additional software can
be used to calculate the mean histogram value from multiple images [132]
[137]. Difficulties in deriving a strain elastography image in EUS result from
very small lesions, deeply located lesions, and pulsations from the heart or large
vessels. Limitations regarding the differentiation between benign and malignant
lesions are caused by fibrosis (lymph nodes and pancreas) [136]
[137]
[138]
[139]
[145], tumor necrosis, and liquid and semiliquid components. A lymph node or
pancreatic lesion must be assessed in the overall evaluation of sonomorphology,
color duplex imaging, and contrast-enhanced harmonic EUS together with strain
elastography. Elastography in EUS is quite helpful for making clinical decisions.
Lymph nodes can be selected for EUS-FNA using strain elastography. A pancreatic
lesion that is hypervascularized and endosonographically soft is highly unlikely to
correspond to ductal adenocarcinoma [146].
Current terminology and guidelines on the use of ultrasound elastography
Current terminology and guidelines on the use of ultrasound elastography
The first European Federation of Societies for Ultrasound in Medicine and Biology
(EFSUMB) guidelines and recommendations on the clinical use of elastography
described the basic principles and technology and clinical applications in detail
[147]
[148]
[149]. This was followed by the
publication of the World Federation for Ultrasound in Medicine and Biology (WFUMB)
2015 and the Society of Radiologists in Ultrasound (SRU) 2015 [150]. All of these guidelines list the relative
advantages of the different elastographic methods. The 2013 EFSUMB guidelines and
recommendations on the clinical use of elastography were updated in 2017, with a
focus on the assessment of diffuse liver disease [103]
[105]. An update was also
provided by WFUMB for liver elastography [80]
and by the SRU [151]. In addition to the 2013
EFSUMB guidelines, updates for non-hepatic applications were published in 2019 [70].
The basic principles of elastography remain unchanged since they were outlined in the
first part of the original EFSUMB guidelines and WFUMB guidelines on this subject
[147]
[148]. The term elastography is considered a type of remote
palpation that allows measurement and display of biomechanical properties associated
with the elastic restoring forces in the tissue that act against shear deformation.
This view unifies the different types of elastography, namely strain imaging,
acoustic radiation force impulse imaging, and shear wave elastography (SWE) and
explains why they all display images with contrast for the same underlying
information, associated with the shear elastic modulus. The type of elastography
most suited for the assessment of diffuse liver disease is SWE, which measures the
speed of a shear wave in the liver. In the 2017 update [103]
[105], VCTE is regarded as a type of SWE, although it differs from other SWE
methods because it uses a body-surface vibration to create a shear wave, which then
travels to the liver, whereas the other methods use acoustic radiation force to
remotely create the shear wave within the liver. VCTE also generates shear waves at
lower frequencies (around 50 Hz) compared to radiation force-based methods (around
200 Hz) leading to different shear wave speed values due to shear wave
dispersion.
A variety of methods that measure shear wave speed should be grouped under the term
SWE [81]
[147]
[148]. Ultrasound elastography
uses ultrasonic echoes to observe tissue displacement as a function of time and
space after applying a force that is either dynamic (e. g., thumping or vibrating)
or varying so slowly that it is considered “quasi-static” (e. g., probe palpation).
The displacement measurements may be represented in an elasticity image
(elastogram), or as a local measurement, in one of three ways:
-
Tissue displacement may be displayed directly, as in the method known as
acoustic radiation force impulse (ARFI) imaging,
-
Tissue strain may be calculated from the spatial gradient of displacement and
displayed, producing what is termed strain elastography (SE), or
-
When the force is dynamic only, the time-varying displacement data may be
used to record at various positions the arrival times of propagating shear
waves. These are used to calculate either regional values of shear wave
speed (without making images) using methods referred to as transient
elastography (TE, including VCTE) and point shear wave elastography (pSWE),
or images of shear wave speed using methods referred to as 2D (or 3D) shear
wave elastography (SWE). All of the methods in this third group come under
the heading of SWE [80]
[103]
[105].
In order to properly introduce the method of elastography to the general public, the
corresponding “How to perform” recommendations are available in addition to the
guidelines [152]
[153]
[154].
Conclusions and future perspectives
Conclusions and future perspectives
Compared to more than 2000 years of palpation, it is possible to look back on the
successful and effective development of elastography thanks to the effective work of
a large number of teams in the last 35 years. Elastography is an important
additional imaging modality in many specialties. It should be actively used by
investigators. In the evaluation of liver fibrosis, histology can be omitted in a
large number of patients. In the overall context with B-mode sonography, color
Doppler imaging, and contrast-enhanced US or EUS, lesions and their status can be
more reliably delineated. Developments that compensate for the limitations of
elastography, e. g., increased fat content, would be desirable to more accurately
predict fibrosis even in hepatic steatosis. More accurate tissue analysis would be
helpful to further delineate focal lesions. In contrast to the modern technique,
there is still a widespread perception that elastography is mainly nice and colorful
but not very helpful. Through continuing education and communication of information,
this misconception should be eliminated so that examiners are even better able to
make accurate diagnoses.
Bibliographical Record
Christoph F. Dietrich, Yi Dong, Xin-Wu Cui, Mathias Fink, Christian Jenssen, Kathleen Moeller, Laurent Sandrin, Sugimoto Tsuneyoshi, Mickael Tanter. Ultrasound elastography: a brief clinical history of an evolving
technique. Ultrasound Int Open 2024; 10: a23786926.
DOI: 10.1055/a-2378-6926
