U

AUUP

BVJa

UataualtRapeimtac

MMn(roC

DD

P

4

ltrasound and Regional Anesthesia

rtifacts and Pitfall Errors Associated Withltrasound-Guided Regional Anesthesia. Part I:nderstanding the Basic Principles of Ultrasoundhysics and Machine Operations

rian D. Sites, M.D., Richard Brull, M.D., F.R.C.P.C.,incent W. S. Chan, M.D., F.R.C.P.C., Brian C. Spence, M.D.,ohn Gallagher, M.D., Michael L. Beach, M.D., Ph.D., Vincent R. Sites, M.D.,

nd Gregg S. Hartman, M.D.

Ua

athrtlo(ippsimw

(

w1cbpmtttcars

ltrasound guidance in regional anesthesia hasgrown in popularity over the past 5 years. Its

ttractiveness stems from the unprecedented abilityo visualize the target nerve, approaching needle,nd the real-time spread of local anesthetic.1 Asltrasound experience grows within the regionalnesthesia community, the limitations and chal-enges begin to declare themselves. Chief amonghese limitations are ultrasound-generated artifacts.ecognition of such optical events combined withn appreciation of the mechanisms involved sup-orts a high quality ultrasound-guided regional an-sthesia practice. The objective of this article (Part I)s to describe the physical properties of ultrasound

ost relevant to the regional anesthesiologist sohat clinical sonographic imaging can be optimizednd common ultrasound-generated artifacts (dis-ussed in more detail in Part II2) can be recognized.

From the Departments of Anesthesiology (B.D.S., B.C.S., J.G.,.L.B., G.S.H.) and Orthopedic Surgery (B.D.S.), Dartmouthedical School, Dartmouth-Hitchcock Medical Center, Leba-

on, NH; Department of Anesthesiology and Pain MedicineR.B., V.W.S.C.), Toronto Western Hospital, University of To-onto, Toronto, Ontario, Canada; and the Department of Radi-logy (V.R.S.), Tufts University School of Medicine, Laheylinic, Burlington, MA.Accepted for publication May 18, 2007.Reprint requests: Brian D. Sites, M.D., Regional Anesthesia,artmouth-Hitchcock Medical Center, One Medical Centerrive, Lebanon, NH 03756. E-mail: brian.sites@hitchcock.org© 2007 by the American Society of Regional Anesthesia and

ain Medicine.

o1098-7339/07/3205-0010$32.00/0doi:10.1016/j.rapm.2007.05.005

12 Regional Anesthesia and Pain Medicine, Vol 32, No

ltrasound Generation, Frequency,nd Wavelength

An ultrasound wave is a form of acoustic energynd is generated when multiple piezoelectric crys-als inside a transducer (i.e., the probe) vibrate atigh frequency in response to an alternating cur-ent. The rapid vibration, which is transmitted tohe patient through a conductive gel, propagatesongitudinally into the body as a short, brief seriesf compressions (high pressure) and rarefactionslow pressure). Each ultrasound wave is character-zed by a specific wavelength (distance betweenressure peaks) and frequency (number of pressureeaks per second). The propagation velocity of aound wave (i.e., acoustic velocity) is fairly constantn the human body (c) and is approximately 1,540

eters per second. Therefore, in the human body,e can use the following equation:

1) c � � · f

here � � wavelength, f � frequency, and c �,540 meters per second. In order to generate alinically useful image, the ultrasound waves mustounce off of tissues and return to the probe. Therobe, after emitting the wave, switches to a receiveode. When ultrasound waves return to the probe,

he piezoelectric crystals will vibrate once again,his time transforming the sound energy into elec-rical energy. This process of transmission and re-eption can be repeated over 7,000 times a secondnd, when coupled to computer processing, willesult in the generation of a real-time 2-dimen-ional image that appears seamless.

The degree to which the ultrasound waves reflect

ff of a structure and return to the probe will de-

5 (September–October), 2007: pp 412–418

tsgow

U

(actp(cdsatpab

itpsaw0

dtpbactTtatnt

foosalistu(

Fdttwsptvirwi

Fpann

Ultrasound Physics • Sites et al. 413

ermine the signal intensity on an arbitrary graycale.3 Structures that strongly reflect ultrasoundenerate large signal intensities and appear whiterr hyperechoic. In contrast, hypoechoic structureseakly reflect ultrasound and appear darker.

ltrasound Interactions with Tissues

As the ultrasound waves travel through the bodyFig 1), they are influenced by reflection, refraction,nd attenuation.4 When an ultrasound wave en-ounters a boundary between 2 different types ofissues, part of the acoustic energy is reflected andart is transmitted. A large and smooth reflectore.g., the needle) acts like a mirror and is hencealled a specular reflector. An irregular surface ran-omly scatters ultrasound and is referred to as acattering reflector. Most neural images are gener-ted based on scattering rather than specular reflec-ion. The amount of ultrasound that is reflected isroportional to the difference in acoustic imped-nce (tendency to resist the passage of ultrasound)

ig 1. The many responses that an ultrasound wave pro-uces when traveling through tissue. (a) Scatter reflection:he ultrasound wave is deflected in several random direc-ions both to and away from the probe. Scattering occursith small or irregular objects. (b) Transmission: the ultra-

ound wave continues through the tissue away from therobe. (c) Refraction: when an ultrasound wave contactshe interface between 2 media with different propagationelocities, the ultrasound wave is refracted (bent) depend-ng upon the difference in velocities. (d) Specular reflection:eflection from a large, smooth object (such as the needle)hich returns the ultrasound wave toward the probe when

t is perpendicular to the ultrasound beam.

etween adjacent tissues. The greater the mismatch (

n acoustic impedance between 2 tissue interfaces,he more energy is reflected back towards therobe, resulting in 2 distinct images on the ultra-ound screen. A bone/soft tissue interface, for ex-mple, reflects 43% of the incoming ultrasoundaves. In contrast, a muscle/blood interface reflects.1% of the ultrasound waves.Clinical pearls: The regional anesthesiologist is at a

istinct advantage when the target nerve is surrounded byissue that has a different acoustic impedance. For exam-le, the sciatic nerve in the popliteal fossa is surroundedy adipose tissue. The large difference in acoustic imped-nce between the sciatic nerve and the adipose tissueauses the nerve to appear clearly hyperechoic relative tohe hypoechoic surrounding adipose tissue (Figs 2 and 3).he reverse is true for the roots of the brachial plexus in

he interscalene region. Here, there is a large difference incoustic impedance between the fascial layers that envelophe plexus and the nerves themselves thereby causing theerves to appear unmistakably hypoechoic relative to

heir surroundings (Fig 4).When ultrasound passes through a tissue inter-

ace (nonreflected), it will likely change its directionf travel. This is a process known as refraction andccurs when the wave reaches a boundary thateparates 2 tissues with different, however slight,coustic velocities. Light also is refracted (Snell’saw) and is the reason a fork appears bent when its inserted into a glass of water. Refracted ultra-ound may not contribute to successful imaging ofhe target structure if a significant amount of theltrasound does not return to the probe. Refractionas well as reflection away from the probe) occurs

ig 2. The short axis view of the sciatic nerve in theopliteal fossa. The large arrow indicates the nerve. Thedipose tissue creates a distinct interface with the sciaticerve which allows for an easy visual distinction betweenerve and surrounding muscle. The nerve is hyperechoic

white) and the fat is hypoechoic (dark).

tto

tmwtdburpvtn

eiitamaditdmcctwts

os

natFieaioacpco

R

amao

sarltt((

Fpbta

Fpts

414 Regional Anesthesia and Pain Medicine Vol. 32 No. 5 September–October 2007

o a larger extent when the angle of incidence be-ween the ultrasound beam and the structure isther than perpendicular.3

Clinical pearls: With respect to needle visualization,he goal of the anesthesiologist is to simultaneously

inimize refraction and maximize reflection back to-ard the probe by keeping the needle perpendicular to

he ultrasound beam as indicated in Figure 5A. Witheeper nerve targets, the angle of incidence between theeam and needle becomes more parallel such that moreltrasound waves are redirected (by refraction andeflection) and fewer waves successfully return to therobe. The end result is that the needle becomes lessisible (Fig 5B). For this reason, many providers preferhe out-of plane needle approach for deeper targeterves (Fig 6).Attenuation is the progressive loss of acoustic

nergy as a wave passes through tissue.4 This resultsn a progressive decrease in the returning signalntensity as the ultrasound travels deeper into aissue bed. The major source of ultrasound attenu-tion is the conversion of some of the acoustic (i.e.,echanical) energy into heat by a process known as

bsorption. Attenuation is directly related to theepth of beam penetration, the type of tissue beingmaged, and varies indirectly with the frequency ofhe ultrasound waves. Different tissues will result inifferent degrees of attenuation. Attenuation iseasured in decibels per centimeter of tissue (dB

m�1) and is represented by the attenuation coeffi-ient of the specific tissue. The higher the attenua-ion coefficient, the more attenuated the ultrasoundaves are by the specified tissue. Examples of at-

enuation coefficients of different physiologic tis-

ig 3. The short axis view of the sciatic nerve in theopliteal fossa. This image comes from an athlete. Notehe difficulty in distinguishing the sciatic nerve from theurrounding muscle. N, sciatic nerve; M, muscle.

ues are listed in Table 1. Figure 7 shows the impact s

f frequency and depth on the attenuation of ultra-ound.

Clinical pearls: While attenuation can have a profoundegative impact on image quality, there are 2 importantdjustments that can be made on the ultrasound machinehat help to overcome some of the effects of attenuation.irst, most machines allow the operator to artificially

ncrease (or decrease) the signal intensity of the returningchoes from all points in the displayed field. This isccomplished by adjusting the gain control higher toncrease the overall brightness. Second, most machinesffer the operator the ability to control gain independentlyt specified depth intervals. This is known as time gainompensation. The time gain compensation should berogressively increased as the depth of penetration in-reases in order to compensate for the corresponding lossf signal intensity (Fig 8).

esolution

Resolution refers to the ultrasound machine’sbility to distinguish one object from another.5 Theost important types of resolution for the regional

nesthesiologist are axial, lateral, and temporal res-lution.Axial resolution refers to the machine’s ability to

eparate 2 structures lying at different depths, par-llel to the direction of the ultrasound beam. Axialesolution is roughly equal to one half of the pulseength. If the distance between 2 objects is greaterhan one half of the length of the ultrasound pulse,hen the structures will appear as 2 separate objectsFig 9). It follows then that higher frequency probesshorter pulse lengths) produce the best axial reso-

ig 4. The short axis view of the interscalene brachiallexus. The arrows indicate the individual roots of therachial plexus. Note the distinct hypoechoic nature ofhe roots in contrast to the hyperechoic fascial sheath. AS,nterior scalene muscle; CA, carotid artery; MS, middle

calene muscle; SCM, sternocleidomastoid muscle.

lqai

8se

FiutN e are md

F(Tdoo

Ultrasound Physics • Sites et al. 415

ution. However, as described above, higher fre-uency ultrasound waves are more readily attenu-ted than lower frequency sound waves, resultingn poor tissue penetration.

ig 5. (A) Long axis image of an 18-gauge needle insertnserted perpendicular to the ultrasound beam, it acts altrasound returning to the probe. (B) The angle of needhereby increasing refraction and reflection away from thote also the reverberation artifact in (A), in which theriscussion of this artifact can be found in Part II.2

ig 6. (A) The in-plane approach for needle insertion. (BC) The out-of-plane approach for needle insertion. (D) The out-of-plane approach has the disadvantage of onlyeep blocks such as the transgluteal sciatic block, the out-f reflection toward the probe and minimization of refra

f the needle to the beam even at extreme depths.

Clinical pearls: High frequency transducer probes (e.g.,-12 MHz) afford high axial resolution of superficialtructures (e.g., axillary region) but have low tissue pen-tration. Low frequency probes (e.g, 4-7 MHz) allow for

plane with the ultrasound beam. Because the needle isrong specular reflector, resulting in a large amount ofrtion was changed from perpendicular to more parallel,e. This is the reason the image of the needle is degraded.ultiple needles visualized under the actual needle. Full

corresponding needle image for the in-plane approach.responding needle image for the out-of-plane approach.

izing a portion of the needle on short axis. However, forne approach may be preferred secondary to optimizationThis is secondary to the near perpendicular relationship

ed in-s a stle insee prob

) Thehe corvisualof-plaction.

deiobppfout

cobacs3ie

pmwsbqtr

ieltii

poucd(

FpwlaTm

FffwmccDttti(gsBtp

416 Regional Anesthesia and Pain Medicine Vol. 32 No. 5 September–October 2007

eeper tissue penetration (e.g., subgluteal region) at thexpense of fine axial resolution. Therefore, probe selections always a trade-off between axial resolution and depthf penetration. When performing a peripheral nervelock, choose the probe and settings with the highestossible frequency that will still afford adequate depthenetration for imaging of the target nerve. See Figure 8or an example of an ultrasound interface that allows theperator to control the frequency (wavelength) of theltrasound. Most ultrasound systems allow the operator

o change through multiple frequencies for a given probe.Lateral resolution (Fig 10) refers to the ma-

hine’s ability to distinguish 2 objects lying besidene another, perpendicular to the ultrasoundeam.5 Lateral resolution is always worse thanxial resolution, thus contributing to more clini-al challenges. Despite the generated 2-dimen-ional image, modern ultrasound machines emit a-dimensional ultrasound beam that diverges as

t propagates through the body (Fig 11). Whenlectronically launched in various sequences and

ig 7. Attenuation. Attenuation is estimated as � � f �ath length, where f is the frequency of the ultrasoundave and � is the attenuation coefficient. Notice the

ower frequency wave (2.5 MHz) has less attenuation atgiven distance when compared with the 10 MHz wave.hus, the 2.5 MHz wave is able to penetrate the tissue

Table 1. Attenuation Coefficients (at 1 MHz)

Material dB cm�1

Bone 20Air 12Muscle 1.2Brain 0.9Fat 0.6Blood 0.2Water 0.002

aore effectively than the 10 MHz wave.

atterns, the collective beams generated from theultiple piezoelectric elements in the transducerill produce the 3-dimensional beam. The

horter the distance between 2 adjacent elementeams, the better the lateral resolution. High fre-uency and focused ultrasound beams generatehe narrowest beams, thus maximizing lateralesolution.

Clinical pearls: The focal zone of the ultrasound beam,ndicated on most screen displays, represents the narrow-st part of the beam and should be positioned at the exactevel of the target nerve. See Figure 8 for an example ofhe focus button on an ultrasound machine. The focuscon that is displayed on the ultrasound screen is shownn Figures 4 and 5 of Part II.2

The limitations of temporal resolution may im-act the regional anesthesiologist. Temporal res-lution is directly related to the frame rate of theltrasound machine. The frame rate of a systemharacterizes how quickly an imaging device pro-uces unique consecutive images called framesin our case an image of needle, nerve, and local

ig 8. An image of a typical ultrasound interface. (1) Proberequency control. In the depicted system and probe, therequency can be adjusted from 3 MHz to 12 MHz. Theavelength can not be adjusted independently; however,anual adjustments to frequency result in corresponding

hanges in wavelength. (2) Overall gain button. This dialhanges how bright or dark the entire image appears. (3)epth control. The objective is to set the depth to just below

he target of interest, thereby optimizing temporal resolu-ion. (4) Focus button. It is important to position the focus ofhe ultrasound beam at the same level as the target ofnterest. This will optimize both lateral and axial resolution.5) Time gain compensation. These toggle dials control theain at consecutive depth intervals. The top dials control theuperficial gain and the bottom dials control deeper gain.ecause attenuation occurs more with deeper imaging, the

ypical pattern of the time gain compensation dials is arogressive increase in gain as indicated in this figure.

nesthesia). High frame rates are critical in cardiac

uAbmtirtuuipfires(ccf

co

gditt

C

aDiu

Firwstoatwtsatl(sphtt

Fhtstbwsiti

Fzazlefd

Ultrasound Physics • Sites et al. 417

ltrasound because of the rapid motion of the heart.s the frame rate decreases, motion-related eventsecome progressively blurred. During nerve blocks,otion occurs with probe movement, needle inser-

ion, and injection of local anesthesia. Therefore, dur-ng these critical moments, a low frame rate couldesult in an ambiguous image. The temporal resolu-ion (frame rate) is limited by the sweep speed of theltrasound beam. In turn, the sweep speed of theltrasound beam is limited by the speed of sound

n tissue, as the ultrasound from the deepest as-ects of the image must return to the probe be-ore the next pulse is generated in the neighbor-ng beam. The sweep speed can be increased byeducing the number of individual piezoelectriclements that make up the larger global beamector or by decreasing the sector scanning anglefor phased array probes). The first option de-reases the lateral resolution and the second de-reases the image field width, underscoring the

ig 9. Axial resolution is the ability to discern objectsn-line with the axis of the ultrasound beam. The axialesolution of an ultrasound wave is dependent uponavelength (�), frequency (f), and the speed of ultra-

ound in tissue (c). In human tissue c � 1,540 m/sec, and,herefore, � � c/f. Axial resolution is roughly described asne half of the pulse length in mm. (A) A low frequencynd a high frequency pulse (2-cycle pulse which is equalo 2 �) propagating toward 2 rectangular objects. (B) Theaves returning toward the probe following the reflec-

ion off of the objects. The blue arrows depict the ultra-ound pulse traveling toward the two objects and the redrrows depict the ultrasound traveling back toward theransducer. The lower frequency ultrasound has a wave-ength that is larger than the distance between the objectsindicated by the black arrows). Therefore, the returningignal from both objects will overlap, and, therefore, therobe will interpret this signal as a single object. Theigher frequency pulse discerns 2 separate objects becausehe wavelength is much shorter than the distance betweenhe 2 objects and the returning waves will not overlap.

undamental concept that temporal resolution i

annot be increased without a compromise sec-ndary to principles of physics.Clinical pearls: The main maneuver the anesthesiolo-

ist can perform to improve the temporal resolution is toecrease the imaging depth to just below the target(s) ofnterest (Fig 8). Additionally, the injection of local anes-hetic should be slow, so as to minimize high velocityissue movement which can blur the real-time image.

olor Doppler

Doppler technology allows for the identificationnd quantification of blood flow. In essence, theoppler principle states that if an ultrasound pulse

s sent out and strikes moving red blood cells, theltrasound that is reflected back to the probe will

ig 10. Lateral resolution is demonstrated here for aypothetical linear ultrasound transducer. The ability forhe ultrasound machine to correctly display 2 objects aseparate structures depends on the relative distance be-ween individual piezoelectric crystals versus the distanceetween the objects. The top 2 structures in this exampleill be imaged as one structure because each falls within

urrounding crystal beams. The red ovals indicate thendividual piezoelectric crystals. For illustration purposes,his figure represents a fictitious situation in which theres no focal zone or divergence of the ultrasound beam.

ig 11. Characteristics of an ultrasound beam. The focalone is where the ultrasound beam width is narrowestnd demarcates the near zone (Fresnel zone) from the farone (Fraunhofer zone). It is also the area of the bestateral resolution because the beam width is the narrow-st at this location. Once the beam extends beyond theocal zone, lateral resolution begins to deteriorate due toivergence. This figure represents a prototypical electron-

cally-focused ultrasound beam.

heqfvve

(

tdbm

ltnob

S

a

auupretatbtuTqtbltitn

1

2

3

4

5

6

7

FvDbtomqtcpD

418 Regional Anesthesia and Pain Medicine Vol. 32 No. 5 September–October 2007

ave a frequency that is different from the originalmitted frequency (Fig 12). This change in fre-uency is known as the Doppler shift.6 It is thisrequency change that can be used in cardiac andascular applications to calculate both blood flowelocity and blood flow direction.7 The Dopplerquation states that:

2) Frequency shift � (2 · V · F1)(cosine�) ⁄ c

Where V is the velocity of the moving object, Ft ishe transmitted frequency, � is the angle of inci-ence of the ultrasound beam and the direction oflood flow, and c is speed of ultrasound in theedia.Clinical pearls: The most important application of Dopp-

er technology for the regional anesthesiologist is to confirmhe absence of blood flow in anticipated trajectory of theeedle, rather than the quantification of the actual velocityr direction of this flow. Doppler information is complicatedy the frequent occurrence of artifact generation.2

ummary

In summary, in order to optimize clinical imaging

ig 12. The Doppler effect. Doppler is used to measureelocity and directionality of objects. In the human bodyoppler is most commonly used to measure velocity oflood flow. (A) The signal from fluid moving away fromhe probe will return at a lower frequency than theriginal emitted signal. (B) The signal contacting fluidoving towards the probe will return at a higher fre-

uency than the original emitted signal. It is also impor-ant to note that the cosine of 0 degrees is 1 and theosine of 90 degrees is 0. Therefore, as the angle ap-roaches 90 degrees large errors are introduced into theoppler equation (see Doppler formula).

nd to appreciate ultrasound-related pitfall errors

nd artifacts, a solid understanding of the physics ofltrasound is extremely helpful. A 3-dimensionalltrasound beam is generated when many multi-le tiny piezoelectric crystals rapidly vibrate inesponse to an electrical current. This ultrasoundnergy is transmitted through tissue where it isransmitted, reflected, scattered, refracted, andttenuated. Fortunately, some of the reflected ul-rasound returns to the probe to be convertedack to electrical energy. This electrical informa-ion is processed by the system’s computer toltimately generate the 2-dimensional image.he anesthesiologist has the ability to control imageuality and appearance by interfacing with the systemo change the characteristics of the ultrasound that iseing sent out such as the frequency, focus, wave-ength, and frame rate. In a similar fashion, the anes-hesiologist has the ability to control how the return-ng image is processed by adjusting such variables ashe gain and various proprietary post processing tech-ologies.

References

. Gray AT. Ultrasound-guided regional anesthesia. Cur-rent state of the art. Anesthesiology 2006;104:368-373.

. Sites BD, Brull R, Chan VWS, Spence BC, Gallagher J,Beach ML, Sites VR, Abbas S, Hartman GS. Artifactsand pitfall errors associated with ultrasound-guidedregional anesthesia. Part II: A pictorial approach tounderstanding and avoidance. Reg Anesth Pain Med2007;32:419-433.

. Weyman A. Physical principles of ultrasound. In: Prin-ciples and Practice of Echocardiography. 2nd ed. Philadel-phia: Lippincott Williams & Wilkins; 1994:3-25.

. Sutton M, Oldershaw P, Kotler M. Textbook of Echocar-diography and Doppler in Adults and Children. 2nd ed.Cambridge: Blackwell Science; 1996:4-18.

. Fiegenbaum H. Echocardiography. 5th ed. Philadelphia:Lea & Febiger; 1994:1.

. Nanda N. Doppler Echocardiography. 2nd ed. Philadel-phia: Lea & Febiger, 1993:7.

. Perinno A, Reeves S, eds. A Practical Approach to Trans-esophageal Echocardiography. Philadelphia: Lippincott

Williams & Wilkins; 2003:77-92.