AJR 2000; 174:9-15
© American Roentgen Ray Society
A Sound Perspective
Honoring Hebar Robarts, MD and G. P. Girdwood, MD
George R. Leopold1
1
Department of Radiology, UCSD Medical Center, University of California, 200 W.
Arbor Dr., San Diego, CA 92103.
Received July 19, 1999;
accepted after revision July 27, 1999.
Address correspondence to G. R. Leopold.
Introduction
The beginning of a new millennium is a good time to take stock of our
accomplishments. In responding to Dr. Rogers' request for a Perspective
article, I had many considerations. According to World Book
Dictionary, the word "perspective" may be defined several
ways: "a view of things or facts in which they are in the right
relation" and "the effect of distance of events upon the
mind" [1]. In the latter
case, I hope the distance will not cause inaccuracies in reporting, but seeing
events incorrectly in perspective is a definite risk. I shall attempt to
combine both definitions in reviewing the evolution of sonography as an
imaging tool within radiology. I am not as interested in detailing historic
developments as in relating the feelings of one fortunate enough to have
participated from nearly the beginning of clinical relevance. On the occasion
of the diamond jubilee celebration of the Radiological Society of North
America, I was privileged to present and later publish an article
[2] in which I reviewed many of
these historical events and the associated technologic developments of
sonography. This Perspective, in contrast, will try to convey the feeling of
being there. To those who participated with me in the early years, it will
most likely evoke nostalgia. Our younger colleagues will probably question the
sanity of their predecessors.
I confess reticence in accepting the responsibility of writing this
article. As I thought about it, however, compelling reasons to proceed
existed. The first stimulus occurred several months ago when a patient came to
our emergency department complaining of right upper quadrant pain and
jaundice. Abdominal sonography showed gallstones, a thickened gallbladder
wall, intrahepatic bile duct dilatation, a dilated common bile duct, and a
large gallstone within it. I thought the images were exceptionally good, and I
commented to the resident how difficult it would have been 30 years ago to get
this diagnostic information at all. Today we can diagnose noninvasively in 20
minutes on an outpatient basis. The second event happened a week later, when
we found some scratchy sonograms from 20 years ago in another patient's film
jacket. "How on earth did you ever make anything out of these?"
was the question posed to me by my younger colleague. It took all the
composure I could muster to say with a straight face, "Those were the
days of the giants." These events convinced me that some perspective of
sonography's past might be worthwhile.
Beginning
In 1842, at the age of 33, Alfred, Lord Tennyson wrote the following verse
of "Locksley Hall," one of the most prophetic poems in literature
[3]:
For I dipped into the future, far as human eye could see,
Saw the Vision of the world, and all the wonder that would be;
Saw the heavens filled with commerce, argosies of magic sails,
Pilots of the purple twilight, dropping down with costly bales;
Heard the heavens fill with shouting, and there rained a ghastly dew
From the nations' airy navies grappling in the central blue.
Tennyson foresaw commercial airplanes and their military potential. I would
like to claim, like Tennyson, that from my early study of sonography I
imagined it would become a mainline technique for radiology. At the beginning,
I hadn't the faintest idea where this fledgling technique would go. Morton
Meyers [4], in his article
"Science, Creativity, and Serendipity," points out many examples
of scientific discoveries virtually hitting their discoverers on the head. I
consider myself a poster child for that paper.
As indicated in my manuscript of 1990
[2], my introduction to
sonography came in 1965 when I was a first-year resident in radiology. As a
punishment for not showing up at noon conference, Elliott Lasser assigned
responsibility for researching this new instrument to the other truant,
Charles Kerber, and me. Dr. Kerber slipped quietly from the scene and left me
holding the bag. Happily, reviewing the world literature on sonography
consumed no more than one or two evenings.
A-Mode Sonography
Diagnostic sonographic studies were begun well before actual images. These
studies derived from simple A-mode oscilloscope displays showing a single line
of sight of the ultrasound transducer. Returning signals from tissue
interfaces were displayed as horizontal deflections from the baseline; their
strength was indicated by the amplitude of the echoes. The term
"A-mode" stands for amplitude modulation. Because cystic
structures possess no internal interfaces, they contrast sharply with solid
tissue, and this contrast forms the principal use of sonography. My earliest
recollection of sonographic studies is of this kind. A particular example
remains vivid.
Several weeks after acquiring the equipment, I was asked to perform a
barium examination on a very large woman thought to have ascites. The
patient's size precluded any sort of radiography; therefore, I attempted a
sonographic examination. No matter where I placed the transducer on the
patient's abdomen, the characteristic cyst pattern appeared
(Fig. 1). I reasoned that this
pattern made ascites unlikely and predicted a huge cyst. At surgery, a
42-pound mucinous cystadenoma was discovered and removed. I was an instant
success. As a first-year resident, I found myself explaining the basics to an
army of skeptics at medical grand rounds several weeks later (some things
never change). Buoyed by this experience, I launched my writing career by
submitting a case report to a major obstetric journal that rejected it.
Crushed by this experience, I waited several years before I tried again. I
still have the original rejection notice in my files.
A-mode sonography was used for other purposes such as analysis of anatomic
structures that were palpable or could be located using other radiographic
methods. For example, the literature contained many references to the analysis
of renal masses first found on excretory urography and later hunted down with
A-mode. In the late 1960s and early 1970s, it was considered necessary to
puncture these masses to confirm their cystic nature, and specific transducers
with central holes were manufactured for this purpose. Other applications of
A-mode sonography included such diverse procedures as measuring the axial
length of the eyeball, the maximum diameter of an abdominal aortic aneurysm,
and the diameter of the fetal head. As incredible as it seems today, the first
fetal biparietal diameter charts were derived from A-mode measurements
obtained from localization of the fetal head by palpation. In this procedure,
the radiologist, looking for the midline complex to assure correct
orientation, applied the transducer to the maternal abdomen and then recorded
a measurement. I have some personal pain associated with this procedure
because it caused me to overestimate the weight of our second child by
approximately 3 pounds—a subject of family discussion for some time
after her birth.
One of the most important applications for early sonography was based on
recognition of the midline complex of the brain seen through the thin portion
of the temporal bone. By comparing measurements made from the opposite side,
the position of the midline structures could be inferred
(Figs. 2 and
3). Before 1975, no CT existed,
and the only other way to study the brain noninvasively was pineal
calcification localization on unenhanced radiographs of the skull. Although
reliable in experienced hands, echoencephalography was difficult to learn.
Problems also arose when the skull was unusually thick, the midline was
distorted by the disease process, or the actual midline shift was located far
anteriorly or posteriorly. Perhaps the worst feature of being able to perform
this test was that it guaranteed a face-to-face confrontation on Saturday
night with a violent drunk suspected of having a subdural hematoma. Few
sonographers regretted the advent of CT of the head and the disappearance of
midline echoencephalography.
B-Mode Sonography
Although A-mode sonography had clinical usefulness, it also had limitations
for physicians used to dealing with anatomic images. Large spike echoes did
not lend themselves to image formation. The solution involved a different form
of echo display. Instead of the returning echoes being shown as vertical
deflections from the horizontal time base, they were shown simply as dots.
Stronger signals resulted in brighter dots on the oscilloscope. This method of
display became known as B-mode, standing for brightness modulation.
One immediate application of this technique was recording the motion of
cardiac structures. By applying the transducer to the left parasternal
intercostal spaces, distinctive motion of the B-mode dots could be seen as the
reflecting structures changed their distance from the anterior chest wall. If
the display was then moved across the oscilloscope face from bottom to top, a
tracing of this movement could be obtained. The resultant tracing was the
distance from the anterior chest wall plotted against elapsed time. At first,
the recordings obtained during the sweep were simply time-exposure photographs
of the oscilloscope face. Later, addition of the strip chart recorder
permitted longer periods of observation. For the first few years, the display
read from bottom to top, but with the addition of physiologic criteria, such
as the electrocardiogram, rotating the display 90° clockwise resulted in a
tracing that read from left to right. This simplification was a great relief
to early investigators who had already begun to develop a wry neck from
prolonged reading in the older manner. These tracings were termed M-mode
(motion mode). With this technique, the diagnosis of pericardial effusion
became relatively straightforward because the single echo of pericardium and
posterior myocardium split into two separate echoes
(Figs. 4 and
5). Of even greater importance,
the motion of the cardiac valves could be recorded and studied in many
conditions. Because the anterior leaflet of the mitral valve remained nearly
parallel to the anterior chest wall during opening and closing, it was by far
the easiest to see (Fig. 6)s. Rheumatic heart disease was still quite common, and the heavily calcified,
slow moving anterior leaflet of the mitral valve in this disorder was a
perfect target (Fig. 7). By
drawing a tangent to this tracing, one could calculate its velocity at any
given time. Studies done after commissurotomy usually revealed changes
indicating the success of the procedure.
View larger version (111K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5. —Rheumatic heart disease and pericardial effusion in 28-year-old
patient. M-mode sonogram shows splitting of posterior heart wall complex in
moving myocardial and stationary pericardial components.
|
|
Many cardiac disorders became the domain of the echocardiographer. One of
the most notable was idiopathic hypertrophic subaortic stenosis. The key
feature of this disorder detectable by echocardiography, in addition to the
thickened interventricular septum, was the systolic anterior movement of the
anterior leaflet of the mitral valve. Systolic anterior movement was often
elusive and had to be sought by provocative maneuvers. The most common was the
inhalation of amyl nitrite, which induced intense peripheral vasodilatation
and forced the heart to work harder. Although this treatment frequently had
the desired effect, it lent a lasting aroma to the lab. I became aware of this
odor one day when I heard our receptionist direct a patient "down the
hall until you smell bananas."
Imaging with B-Mode
Although radiologists were heavily involved in early echocardiography, most
remained unconvinced about the "still newer kind of ray" until
cross-sectional images became possible. Joseph Holmes
[5] and Douglas Howry of the
University of Colorado used novel means of applying the ultrasound beam to the
patient. Perhaps the most famous was to place the patient in a water bath.
Then using the B-mode technique, the radiologist would rotate the transducer
around the patient. The tiny flickers of light produced on the oscilloscope
were captured by time-exposure photography for later inspection. These images,
derivatives of B-mode, were termed "B-scans." Although totally
unacceptable for examining critically ill patients, early B-scans produced
some surprisingly good cross-sectional images.
Engineering progress soon brought the transducer out of the water bath and
placed it at the end of a three-rodded arm with potentiometers at each angle
(Fig. 8). These scanner arms
and potentiometers were connected with wires that required almost constant
recalibration to assure positional accuracy. When recalibration was not done,
truly strange results occurred. Fetuses with rectangular heads were common
during this period. The wires were completely exposed and susceptible to all
sorts of substances and disasters. The greatest trauma occurred from
mechanical friction as wires passed over the pulleys of the arm. Eventually a
wire would break, make a loud snap, and spew pieces across the laboratory.
Although these noises were frightening enough to those accustomed to this
event, patients' reactions were especially remarkable. On more than one
occasion, I saw patients decide they had experienced enough of the new
diagnostic method and bolt from the room.
View larger version (124K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 8. —Early articulated arm, contact B-scanner. On left, transducer hangs
from end of three-rodded arm. On right is recording console, which has both
large recording phosphor and camera for direct photography.
|
|
The new scanner arms eliminated the need for a water bath, but to assure
adequate contact with a patient's skin, a variety of messy substances were
used. The net result was that no pioneers could be found with unstained
clothing. The pioneers always wore bow ties—not as a fashion statement,
but to avoid continuous tie soiling. A corollary occurred among early
sonography salespeople. Because midline echoencephalography was such an
important application, many salespeople found it useful to have a small dab of
acoustic gel behind the ear for instant demonstration purposes in the
radiologist's office.
After suitably coating the patient (and usually a portion of oneself) with
one of the messy substances, the sonographer then placed the transducer on the
patient and mechanically sectored it back and forth. The construction of the
arm kept it locked in a preselected plane. By insonating from many different
angles within the plane, one hoped to obtain a complete collection of internal
reflections. A single cross-sectional scan usually took from 10 to 20 sec. The
returning echoes were amalgamated in a final image in a clever way. The face
of the recording oscilloscope was coated with a phosphor to which the small
flashes of light would stick. At the completion of the scan, the final image
was either photographed or translated to heat-sensitive paper. This chain of
events came with a number of annoying side effects. When it was new, the
phosphor was an excellent medium to retain light flashes, but repeated use led
to a deterioration of its performance. To make matters worse, this
deterioration was more severe in the center of the screen, where virtually all
imaging was performed. A situation was created in which even weak echoes were
recorded at the periphery of the image while closer to the center strong
echoes might not "stick." Further difficulties came from the
bleeding of strong echoes in the phosphor, somewhat analogous to the blooming
artifact seen with today's color Doppler sonography
(Fig. 9).
The permanent recording medium was even more problematic. Because this era
was the beginning of instant photography, it seemed a natural environment for
sonograms. Each study, consisting of 20-30 images, was duly photographed for
archiving. Early film, however, had the nasty habit of fading or disappearing
after a few years. Squeegeelike devices were used to put a protective film on
instant pictures. So it was in sonography—day after day. Although most
of us were convinced that ultrasonic energy in the diagnostic range was
harmless, we also believed that the noxious coating dissolved fingertips
effectively. The finished product was a string of images taped together with
masking tape. These could then be folded like a chain of picture postcards and
fitted neatly in a shirt pocket. Whipping out a packet of these for colleagues
to inspect often added to the carnivallike aura of the early sonographic
laboratory.
Photographic film was later replaced by heat-sensitive paper—usually
large sheets of paper requiring more storage space. Although the sheets
required no preservative coating, they too deteriorated over time and were
never totally satisfactory. The introduction of the multiformat camera was one
of the most significant advances in our discipline because images were now
recorded on X-ray film. This advance greatly facilitated storage but had an
unexpected benefit. Because sonographic studies were now on X-ray film, they
tended to be stored with the rest of the patient's images. Radiologists who up
to that time chose to ignore these early "weather maps" were now
forced to deal with them. More importantly, because sonograms were now in a
familiar format, many radiologists now pronounced themselves capable of
interpreting them.
The actual sonograms generated by these early machines required
considerable interpretative skills (then referred to as the reading of tea
leaves by our chief of urology). Because only the strongest echoes, usually at
the boundary of organs, could be recorded, these scans were referred to as
"bistable," meaning either the signal was or it wasn't there. I
distinctly remember running around the halls of Presbyterian Hospital in
Pittsburgh with images of a patient with autosomal dominant polycystic kidney
disease (Fig. 10) that I
forced on anyone who would look at them. Many people suggested that these
scans looked like Rorschach tests. Other less diplomatic souls suggested that
I should consider seeing a physician who dealt with Rorschach tests. This
skepticism was not unique to medicine in the eastern United States. Shortly
after moving to California at the completion of my residency, I was introduced
at a party to a famous Los Angeles radiologist, who shall remain nameless.
When I told him of my interests, he said that he didn't have time for gadgets
and promptly walked away from me.
In addition, orienting the sonograms was a problem. Because no conventions
existed regarding cross-sectional images, we simply made up our own. Some
sonographers, including me, chose to orient sonograms as if looking from the
top down. In this scheme, the liver would be on the right when viewed in the
axial plane (Figs. 11A and
11B). We told everyone that
this approach was chosen because of the cerebral nature of people pursuing
this nascent technique. Actually, the reason was that the same convention was
found in an ancient cross-sectional atlas, first published in 1911 by
Eycleshymer and Schoemaker [6].
This remarkable book had been used for many years by radiation therapists in
planning and calculating patient dosage. The premise was to assume everyone's
internal organs were in exactly the same place and relationship to one
another. To make matters worse, the cadaver used in the atlas was quite obese.
Early sonographers, not realizing the best patients for sonography were thin,
wasted time trying to recognize anatomy.
View larger version (95K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 11. —Healthy patient. [In 1971, convention was to view sonogram downward
from above.]
B, Sagittal B-mode sonogram at level of gallbladder. Note
"bistable" nature of image—all echoes are of same
intensity.
|
|
When radiologists and obstetricians began to show their displays in reverse
format, this change caused considerable confusion at national meetings when
combinations of the two formats were used. The debate was finally settled in
favor of the gynecologic approach because of the development of CT of the head
and a consuming desire to standardize. Nevertheless, on that day in our
laboratory when the liver leapt from one side of the abdomen to the other,
chaos reigned supreme and continued for several weeks. It was not a great
inconvenience to referring physicians, however, because for them
interpretation was equally mysterious in either format. At the same time,
whether one should show the images as black on white or white on black was
debated. Many scientific and nonscientific arguments were advanced to support
both positions. The opinion that it was far easier to see stars at night
(white on black) than it was during the day eventually resolved the
debate.
We spent our days hoping for thin patients with fluid-filled disorders. It
would be several years before the gallbladder could be seen on a routine
basis. In one of the first textbooks on sonography, I stated that sonography
would probably not help gallbladder disease
[7]. Without question, the most
significant event in the history of sonography occurred in 1973 with the
advent of the gray-scale technique. Before scan converters, many sonographers
suspected parenchymal echoes did exist on the basis of a seldom-used method
called the open-shutter technique. In this process, the phosphor-coated
oscilloscope was bypassed, and the returning flashes of light were allowed to
strike the photographic film directly while the shutter was held open. This
procedure produced a weak gray-scale effect and pointed the way to future
investigation (Fig. 12). The
technique was problematic because the operator was completely unaware of what
was occurring during the scan until the film was developed. When the scan
converter appeared, this technique rapidly went the way of dinosaurs.
The introduction of analog scan converters in our machines was like lifting
a great veil from the sonograms (Figs.
13, 14,
15,
16). For the first time, we
were able to confirm that internal organs had parenchymal echoes.
Unfortunately, no one knew what the internal patterns were supposed to look
like and, because the number of controls accessible to the operator was large,
the results were variable. To make matters worse, the scan converters were
highly unstable. At their best, the scans produced exquisite images of the
internal anatomy. The solution came with the introduction of the digital scan
converter, which was intrinsically much more stable. Nevertheless, some
radiologists claimed that their old analog units produced better pictures. The
weight of evidence, however, soon suggested otherwise.
The appearance in 1975 of primitive realtime sonographic units spawned a
new series of arguments among early sonographers. The degraded image of early
linear arrays led many to believe the new stepchild would never achieve
maturity. In addition, frame rates were pitifully slow, often below the eye's
flicker fusion frequency. The result was a steadily blinking image that to
many seemed capable of initiating an epileptic seizure or at least a giant
headache. The functionalists among us were willing to overlook these
deficiencies whereas others, including me, clung to static images much longer.
Francis Weill [8], a true
sonographic pioneer in France and a major advocate for real-time sonography,
refers to me in the first chapter of his book as a "salami
slicer." (On the occasion of his retirement, I participated in an
international videotape gift to him of me in my kitchen slicing salami.)
Happily, this dilemma was solved by improvements in real-time technology that
rendered real-time sonograms equivalent in resolution to static images.
Devices using phased array transducers subsequently advanced the realtime
concept even further.
After these improvements were made, the additions of Doppler sonography
techniques, sonographic contrast material, and three-dimensional imaging have
all contributed to sonography's advancement. With this explosion of
technology, sonography has changed. No longer do most sonographers know one
another. The nostalgia is more than offset by the gains made in improved
patient care.
Conclusion
I hope the foregoing account has been a pleasant reminder to older
radiologists and an informative narration to younger ones. This record is not
intended to be all-inclusive, but rather to represent impressions of one
fortunate enough to have been a front-row observer. Certainly sonography is
now thriving and is considered mainline by most radiologists. To see
sonography evolve from a laboratory curiosity to an accepted technique in
three decades is truly gratifying. The journey has been most remarkable and
rewarding. Sonography's phenomenal growth is a testament to people with many
different skills: engineers, manufacturers, technologists, and physicians.
Without all these groups, sonography would have failed. Supportive mentors are
important. Teachers, colleagues, and patients have encouraged me every step of
the way. It doesn't hurt to have a little good luck, either.
References
-
Barnhart C, Barnhart R. World book dictionary. New
York: Doubleday, 1979
-
Leopold G. The Radiological Society of North America: diamond
jubilee lecture—seeing with sound. Radiology
1990;175:23-27[Abstract/Free Full Text]
-
Tennyson A. Locksley hall. In: The literature of
England, 3rd ed., Chicago: Scott, Foresman,
1948
-
Meyers M. Science, creativity, and serendipity. AJR
1995;165:755-764[Abstract/Free Full Text]
-
Holmes J. Early diagnostic sonography. J Ultrasound
Med
1983;2:33-43
-
Eycleshymer A, Schoemaker D. A cross section atlas of
anatomy, 2nd ed. New York: Appleton Century Crofts,
1970
-
Leopold G. Abdominal ultrasonography. In: King D, ed.
Diagnostic ultrasound. St. Louis: Mosby,
1974:260-272
-
Weill F. Ultrasound diagnosis of digestive diseases,
3rd ed. Berlin: Springer Verlag,
1990:3
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
|
|
 
C. R. B. Merritt
Sonography: Unlikely Success, Uncertain Future
Am. J. Roentgenol.,
May 1, 2000;
174(5):
1201 - 1202.
[Full Text]
[PDF]
|
|
|
|
|
|
|
L. F. Rogers
Celebrating the ARRS Centennial
Am. J. Roentgenol.,
January 1, 2000;
174(1):
1 - 1.
[Full Text]
[PDF]
|
|
|
Top
Introduction
Beginning
