Properties of Sound

Hooke’s law: F = -kx

wave propagation in simple harmonic motion (phase lag between particles), displacements vary sinusoidally; medium must be elastically deformable

transverse waves =- particle displacement perpendicular to direction of propagation

longitudinal/compressional/pressure waves = displacement parallel to propagation with increase/decreases in pressure with compressions and rarefactions (eg sound with larger changes in pressure = loud)

T = period = time for one oscillation; f = frequency; λ = wavelength; v = velocity

v = λ/T = fλ; dependent on density and compressibility of medium (ie mass, particle spacing, strength of interactions; reduces if highly compressible eg air)

human hearing 20-20,000Hz, ultrasound 2-15MHz

v in air = 330m/s; liquid = 1480; soft tissue = 1540; bone = 4000; ΔvΔλ

A = amplitude = (max) pressure amplitude; (max) particle displacement; or (max) particle velocity

W = power (W or mW, watt)= rate of energy through whole area of beam, not constant w time and usually refers to average power

I = intensity (W/cm2 or mW/cm2)= energy through unit area in unit time A2; not uniform in beam

dB = decibel = relative change in intensity, amplitude or pressure = 10log10(I/I0) or 20log20(A/A0); 3dB is a double in I

Z = acoustic impedance (rayl = kgm-2s-1) = resistance = ρv; ρ = density


principle of superposition = total displacement is the algebraic sum; destructiveless/no particle movement; constructivegreater movement; wave continues unimpeded


divergence of wave when from a small source or aperture, greater when source/aperture similar to λ

resonance and damping

applying forces at the resonant frequency (f0) results in increased A and continues unless damping force applied

damping reduces duration of resonance and increases range of frequencies present = bandwidth Δf (difference between freq above and below resonance freq at which intensity is reduced to half)

Δt (duration of damping) = 1/Δf

ring down-time = interval betw initiation of wave and complete cessation of vibrations

Q-factor = quality factor = f0/Δf; high Q have narrow BW, but longer SPL

Interaction with Matter

insonation = striking of tissue with sound

attenuation = exponential reduction in A or I due to absorption (major factor in solids; almost sole factor in liquids), scatter, reflection, beam divergence and refraction

μ = attenuation coefficient (dB/cm), 0.5 dBcm-1MHz-1 in soft tissue (air 12, bone 5, muscle 1-3, liver 1, kidney 1, blood 0.2, fat 0.6, water 0) and f; Id/I0 = e-μd = e-μvt

HVL = half value layer = thickness necessary to attenuate intensity by 50% (3dB)

reduction in average f with depth due to preferential attenuation of high f


at interface of different Z, a proportion of energy is reflected (reflection coefficient, R) and rest transmitted; most is reflected at high differences in Z

specular reflection = when dimension of interface λ

R = reflection coefficient = Ireflected/Iincident = which increases with angle of incidence

angle of incidence = angle of reflection

focusing or defocusing can occur at curved interfaces


Snell’s Law: sinθ1/sinθ2 = v1/v2; θ1 and θ2 = angle of incidence and refraction measured from principle line (perpendicular to surface)

when moving into slower speed medium λ reduces and direction changes towards principle line

total internal reflection occurs when v2>v1 when θ > critical angle = θc = v1/v2

curved interfaces may cause defocusing


diffuse reflection when dimension of interface slightly < λ

Rayleigh scattering when interface λ, scatter uniform in all directions f4 (eg RBCs)

scatter ‘signature’ from the characteristic structure of organs


relaxation processes = transfer of energy to random molecular motion (=heat) or internal molecular energy and is f


piezoelectric crystals (eg natural quartz, artificial PZT lead zirconate titanate) have asymmetric charge distribution with molecular dipoles aligning in direction of electric field (up to 600V but usually ~150V, ↑V↑A), causing deformation with return to normal shape with voltage removed; when compressed they cause a potential difference

artificial piezoelectric crystals (synthetic ceramic, usually PZT) made by heating above its Curie temperature (328-365°) and allowing to cool in strong electric field

if transducer heated above Curie temp (T that dipole alignment occurs), alignment will be lost

Transducer Design and Operation

matching layer (coupling medium) 1/4 λ thick with Z between crystal and tissue to cause destructive interference in reflected sound, permitting ~100% transmission

element of PZT crystal, thickness selected for desired f0,= ½ λ causing a standing wave (constructive interference betw reflecting waves)

electrodes plated onto crystal faces

backing/damping block (absorbs backward direction of ultrasound energy, attenuates stray US signals from housing and dampens the transducer vibration)

acoustic insulator and casing/housing for electrical insulation

continuous wave – continuous alternating V at desired f (CW Doppler units), very narrow-band

long pulse – desired length of V applied or crystal lightly damped (pulsed Doppler), narrow-band

short pulse –short V spike <1μs (causing crystal to ring at f0) with heavy damping (for imaging, pulse length few λ = 1μs), broad-band

frequency content of pulse changes with distance due to differential absorption at higher f

Beam Profile and Focussing

dead zone = 1st few mm lost in main bang since crystal needs to recover from emission process

Fresnel zone = near field; length = D2/4λ, parallel beam; D = width of transducer

Fraunhofer zone = far field; sinθ = 1.22λ/D, intensity distribution not uniform due to interference, more uniform with broader bandwidth

side lobes = off-axis peaks from constructive interference at certain angles

grating lobes = off-axis for phased/linear array transducer (acting like diffraction grating)

focal zone = range that beam width is less than twice the width at focus (always within the near field of equivalent unfocussed), and reduces with degree of focusing; divergence distal to this zone is greater than equivalent unfocussed transducer

beam formation – transmit and receive processing used to define the beam for steering and focussing

apodization – weighting function applied to different channels to reduce side or grating lobes

mechanical focusing

internal focusing – curved transducer crystal

external focusing – acoustic lens or concave mirror

usually applied in 2D

double focussing effect – reception and transmission from/to focus is greatest

electronic focussing

multi-element arrays fired in sequence from outer to inner, with phase delays also introduced into returning echoes

linear/phased element arrays are linear, so focussing is only 1D with mechanical focussing perpendicular to scan plane

variable transmit focussing increases firing delays for shorter focal zone

multi-zone focusing – more than one pulse (up to 5) fired for each line using different focal ranges and echoes only recorded from restricted range of depths near the focus; good lateral resolution over extended range at expense of frame rate, but some machines can transmit pulses simultaneously to maintain frame rate

dynamic receive focus – delays added for reception focussing and varies during reception time

Mechanical Sector Scanners

rotating or oscillating transducers in a fluid-filled scanhead, or oscillating acoustic mirror

pulse emission and reception occurs at preset angular displacements, operating only through the preset acoustic window

3 rotating transducers of different frequencies can be used

line density decreases with D

crystal with fixed mechanical focus, or annular array with variable electronic focusing

trapezoidal field if fluid path between crystal and housing is long

Electronic Linear Array

multiple crystal elements arranged in a row (128-192 10x1mm elements)

large footprint, rectangular FOV

elements are fired in groups (larger D→ longer near field; also electronic focussing can be used), with echoes corresponding to middle element of the firing group and moved along array by one element at a time

line density doubled by using odd-even firing sequence (middle element moves by ½ a space)

perpendicular focusing by curved crystal elements or acoustic lens

lateral resolution defined in 2 planes

2D array transducers allow variable electronic focussing in 2 directions

convex/curved linear/curvilinear arrays have a curved face for larger FOV, however line density reduces with D

Electronic Phased-Array Sector Scanners

multi-element array with smaller elements and total length; all elements are fired for each scan line

electronic beam steering produces sector-shaped FOV (up to 90°) by delays in firing of elements

size and number of crystals affect sensitivity at extremes of scanning field and suppression of grating lobes

alteration of firing delays (non-linear sequence) used for focussing

line density decreases with D, and lateral resolution reduces at edges

Mechanical Annular Array Sector Scanners

elements arranged in series of concentric annuli with real-time imaging by mechanical oscillation or rotation of transducer

electronic focussing symmetrical about central axis of beam

sector angle can be >90° with resolution and sensitivity constant at edges

Scanning Techniques

should use the highest freq transducer that can reach target depth; 7.5-10MHz for 1-3cm, 2-3.5MHz for 12-15cm

waterbath scanning places superficial structures beyond the dead-zone into the focal zone; also increases FOV

biopsy probes have a slot for biopsy needle or have a clip-on biopsy guide

intracavity probes may be linear array, sector scanner or rotating transducer to produce complete 360° image

surgical probes may have special shape (to optimize access), are sterilisable (but not by autoclaving) or have a sterile sleeve

Pulse-Echo Imaging

SPW = spatial pulse width = pulse length ≈3λ

PRF = pulse repetition frequency = number of V spikes per second; limited by time for echoes to return ≈ 1000-5000/s

duty cycle = percent of total operation time in emission ~0.1% = PRF x pulse duration

A of echo received determined by A of emitted pulse, distance, μ, ΔZ at interface, type of interface, θ incidence, f

A-Mode (Amplitude Modulation) Imaging

cathode ray oscilloscope displays voltage from transducer crystal (after processing) vs time base; displaying single line of reflections

B-Mode (Brightness Modulation) Imaging

512×512 or 512×640 pixels 8bit (256 shades of grey)

static B-mode scanners

transducer mounted on gantry/tracking system with position and amplitude information passed to scan converter

simple scanning – interrogation of tissue region once only

compound scanning – each region interrogated from several different directions

real-time B-mode scanners

automatic, rapidly repeating scanning action at 15-60 frames/s

extra lines may be interspersed by averaging adjacent lines

max frame rate, Fmax = ; D = depth, N = scan lines per frame, v = 154,000cm/s

freeze frame

temporal frame-averaging (either in real time or with freeze-frame only)

limited FOV and lack of frame of reference

M-Mode = T-M (Time Motion) Imaging

sweeping single line of B-mode echoes across display, with movement causing deviation of echo line

acquisition of B-mode and M-mode simultaneously is possible with electronic rapid switching arrays

time base = rate of sweep can be varied

3D Imaging

1D or annular arrays with positioning device

  • large probes with integrated position-sensing in housing and imaging plane moving automatically
  • transducer mounted on external motorized drive
  • external position-sensing device (does not drive scanning)
  • freehand scanning without position sensing

2D array transducers to acquire the complete volume set

2D array transducers or fast mechanical scanning systems to sweep the volume in <1s to enable real-time performance (4D)

multiplanar reformatting techniques

Doppler Ultrasound

Doppler effect = relative movement causing change if frequency by Doppler shift

Doppler shift equation: Fd = shift frequency = F0-Fr = 2F0(cosθ)u/v; F0/Fr = transmitted and received f; u = velocity of reflector, v = velocity of sound; usually if audible range (<20kHz)

best F0 balance between attenuation (increase w f) and scatter from RBC (f4); requires small BW (long pulses)

Continuous Wave (CW) Doppler

lightly damped or undamped sinusoidal V of desired f with very narrow BW; second crystal for reception

no depth perception and area of sensitivity determined by region of overlap

beat phenomenon – added sounds with different f have beats at Fd

multiplication of signals produces signal with Fd and original f (that can be filtered out)

F0 and amplified Fd signals demodulated and low-pass filtered (filtering out F0), which is then amplified and sent to earphones/speakers or chart recorder

quadrature phase detection permits detection of direction of movement, but flow from multiple vessels may be superimposed; thus spectral analysis may be used

high pass/wall filters eliminate the low frequency components that represent noise, set at 50-100Hz

Pulsed Wave (PW) Doppler

range gating- only echoes from particular depth range (sample volume) processed, determined by pulse length, beam shape and time gating

short bursts of several cycles, low BW (relatively long hence poor depth resolution) repeated until enough data for accurate estimate of Fd

direction detectionDoppler detection (extracts Doppler shift waveform)spectral analyser (differentiates between various f components)output

direction detection

QPD = quadrature phase detection = Fd sent to separate channels and mixed with reference signals from transmitter 90°out of phase; thus 2 signals produced with phase depending on whether Doppler shift is positive or negative; further phase shift and adding/subtracting yields forward flow in one channel and reverse flow in the other

heterodyne detection mixes Doppler signal with reference signal such as 5kHz, so beat frequency >5kHz for forward and <5 for reverse flows

pulsed/spectral Doppler

spectral analysis = Doppler frequency spectrum

duplex systems combine imaging and Doppler with transducer alternating between the two; electronic beam control allows rapid switching and more flexible control of beam direction whereas mechanical sector probes will need to stop and start the motion

brightness-modulated frequency/power spectrum vs time with sufficient N of samples (~ 64-128 pulses) in short time intervals (10ms to ensure stationarity)

FFT separates waveform into frequency components

high-pass/wall/thump filter removes clutter from slow moving reflecting structures

θ must be known to convert f to velocity, when >60° errors in θ cause large errors in cosθ; when <20-30° refraction and aliasing problems can arise

beam steering or wedge betw transducer and skin can be used to reduce angle

laminar flow – fast in middle with reducing velocity peripherally due to friction; in large smooth-walled vessels

blunt flow – uniform in centre, drops off at vessel wall

turbulent flow – spectral broadening downstream from stenosis

high peripheral resistance causes low peak systolic A and reverse/low diastolic flow, possible triphasic waveform

low resistance causes high systolic A with continued relatively high diastolic forward flow

systolic/diastolic ratio = peak systolic freq or velocity (S)/end diastolic freq or velocity (D)

RI = resistive index = (S-D)/S

PI = pulsatility index = (S-D)/mean freq or velocity

vessel flow and resistance influenced by HR, BP, vessel wall length, elasticity and extrinsic compression


sampling theory: for accurate Fd calculation, signal must be sampled at least twice per cycle, hence max Fd measured = ½ PRF = Nyquist limit

aliasing – when frequencies above Nyquist f result in much lower/reverse flow calculated Doppler shift

max PRF is limited by depth

high-PRF mode – 2nd pulse transmitted before 1st pulse returns, however causes range ambiguity as 1st received signal may arise from 2nd pulse coming from shallower region; unambiguous if shallow region is in area of no flow

can also be avoided by using lower Fd by reducing transducer f or increasing θ

colour flow/Doppler/velocity imaging (CFI/CDI)

red and blue represent flow direction, third colour (usually green) may indicate variants of velocity (spectral broadening)

6-18 pulses (instead of 64-128) for each line of data; Doppler estimation with autocorrelation comparing consecutive lines from same position, calculating mean Fd, variance and direction of flow

angle correction hence velocity not possible

requirements for optimum B-mode and Doppler are different; asynchronous system acquires each separately

multiple pulses, hence reduced frame rate; increased by using larger pixels for colour information or reducing colour ROI to smaller portion of FOV

tend to use lower PRF than PW, hence increased aliasing

flash artefact = solid tissue motion

time domain correlation – velocity calculation in CFI by tracking unique echo pattern (speckle) or scatterers

unable to display entire frequency spectrum

susceptible to noise

colour amplitude imaging (CAI) = power Doppler (PD) = colour Doppler energy (CDE) = ultrasound angio

power of Doppler signal used instead of mean Fd for colour-coding, sensitive to presence or absence of flow

autocorrelation used to determine presence of motion

advantages: increased display dynamic range, independent of angle, no aliasing, less susceptible to noise (which has low power, hence able to increase gain), better vessel wall definition

disadvantages: susceptible to flash artefact (reduced with temporal averaging), no information on direction, velocity or pulsatility of flow, more affected by attenuation

Tissue Harmonic Imaging

wave steepening = positive peak of wave has slightly higher speed than negative peak; greatest at high amplitudes

harmonic frequency = distortion equivalent to adding higher frequency components at exact multiples of fundamental frequency, ie peaks at f0, 2f0, 3fo etc

harmonic B-mode images formed from 2f0 (which is formed at most intense part of beam), hence reduces side lobes and thus image clutter improved lateral resolution

harmonic beams not fully developed near transducer, so images are cleaner and free of reverberation artefacts

transmit BW must be narrower so transmit and receive BW are separable and fit within total BW capability of transducer, so requires longer SPW and thus reduced axial resolution

Receiver Functions and Processing

dynamic range (DR, in dB) = range of signal magnitudes that can be handled by system without distortion, >100dB, which becomes compressed and remapped to capabilities of grey-scale monitors (35-60dB)

preamplifier – in transducer, to avoid loss of low level signals in cable

radio-frequency (RF) amplifier

  • amplification = gain (dB) = ratio of output to input power = square of voltage ratio >60dB
  • TGC = time gain compensation – selective amplification of deeper structures to compensate for attenuation varying with e-μvt, usually displayed on logarithmic scale dB vs D ~50-80dB
  • logarithmic signal compression during amplification process as DR of echoes is larger than what storage and display systems can handle

demodulator – signals rectified (given absolute values) and smoothed so each echo represents a single spike only

video amplifier – rejection/suppression/threshold may be used to eliminate pulses below rejection level/threshold for removal of noise

preprocessing – manipulation including TGC, logarithmic compression, write zoom, frame averaging (reduces background noise), further linear or non-linear during digitisation

digital scan converter – stores echo data and converts it into format for display with DR comparable to rest of imaging system; digitisation via ADC 8bit (256:1 V range65536:1 I range48dB DR) or 10 bit (60dB DR)

postprocessing – linear or non-linear conversion of stored values to V with pre-programmed assignment curves (eg low-level expansion/compression), windowing and levelling, read zoom (image magnification), spatial image smoothing, distance measurements

video signal sent to monitor, photographic monitor (high resolution TV monitor from which hard copy is photographed), laser imager or videotape unit

Operator Controls

transducer selection (frequency, diameter, focal zone, format/type)

output power

amplification/overall gain

TGC near gain (min A), delay (depth that near gain applies), slope (dB/cm), far gain (max A); or slide controls to independently set A at regular depth intervals

suppression level

DR (at amplifier or preprocessing stage, low DR increases contrast; usually ~40-60dB)

preprocessing and postprocessing options


monitor brightness and contrast

Image Quality

Spatial Resolution

axial/depth/range/longitudinal resolution depends on SPW ~3-4λ, with resolution limit SPW/2 = 2λ ≈ 0.5mm; hence better for high freq or shorter SPW (more damping); constant with depth

lateral/horizontal resolution (≈2-5mm) determined by beam width and height/thickness (hence transducer diameter, freq, focussing, depth, type of reflector, power output, DR)

elevational/Azimuthal resolution needs to be defined for linear and phased-array transducers and is dependent on transducer height (thickness of image), usually the worst cf lateral and axial


assumptions made about constant speed, echoes from central axis, straight line propagation and constant rate of attenuation (or correct TGC)

echoes may be displayed at incorrect position, may not be displayed, incorrect amplitude (partial volume effects) or be joined (resolution limitation)

reverberation = multiple reflections, reduced by changing scanning angle to avoid parallel reflectors)

  • requires two strongly reflecting interfaces (one usually transducer-skin) causing multiple reflections
  • diffuse reverberation combines scatter in superficial tissues layers resulting in echoes in superficial portion of fluid structure
  • comet tail – multiple short bright bands from two very closely spaced interfaces

multipath reflection – beam reflected off-axis and strikes another reflector before returning to transducer (usually displaying in a region of weak signal)

mis-registration – refraction (eg lens artefact through rectus abdominus)

speed displacement – moving through structure with significantly different sound velocity (causing depth error)

depth aliasing

beam width = partial volume effect

  • echoes from edge of beam displayed in central axis resulting in lines (instead of dots)
  • false echoes brought in from outside plane
  • oblique interfaces appearing falsely thicker

posterior enhancement – overcompensation of TGC deeper to relatively low attenuation structures (fluid-filled cysts)

acoustic shadowing

  • lack of insonation of an area due to highly reflecting interface (air, calculus) or strongly absorbing medium (calculi) and inadequate TGC
  • at curved interface beam may be reflected and refracted with beam diverging or total internal reflection (low-to-high v at shallow angle); most pronounced at edges of round structures (cysts) resulting in narrow shadow

side or grating lobes – side and grating lobes have much lower amplitude but may give rise to echoes at strong reflectors; reduced by changing focal zone, transducer or repositioning

mirror image – high level echoes returning are scattered from proximal tissue before reflecting back from initial interface; proximal tissue displayed distal to reflector

focal zone intensity may be brighter than elsewhere

electronic noise give characteristic fine snowstorm if gain is too high

resolution limitations from separate reflectors appearing as one

incorrect TGC, overall gain, preprocessing, postprocessing may cause underwriting, overwriting or non-uniformity

poor transducer-skin coupling

poor choice of beam path or scanning angles

patient movement

Doppler artefacts – aliasing, flash artefact, attenuation or mirroring, colour bleed or noise in areas of no flow (from high gain), false spectral broadening (from high gain, DR, excessively large sample volume or placement close to vessel wall where velocity is lower)

Quality Control

basic tests/checks include transducer element dropout, transducer cables, keyboard controls, brakes, cleaning, fault reporting

phantoms should have similar acoustic properties to human soft tissue (0.7dBcm-1MHz-1) with non-linearity which determines rate of harmonic generation

AIUM (American Institute of Ultrasound in Medicine) test phantom has fluid bath (10% ethanol in water with v 1540m/s @20°C) crossed by wires, but has very low attenuation and high degree of non-linearity with low backscatter

  • axial and lateral resolution with narrower wire spacing
  • beam profile with length of echo across the beam measure of beam width
  • dead zone – closest wire that can be imaged
  • calibration of measurements
  • registration – placement of echoes in correct anatomical relationships
  • dynamic range

tissue-mimicking phantoms also contain similar attenuation and backscatter properties as tissue; base material having desired sound speed and additional solid particles to produce backscatter

  • TGC operation
  • sensitivity (and SNR) = max D that scatterers can be visualised
  • disuniformity can arise from transducer problems (failed elements), banding disuniformities (if multizone focus)
  • contrast resolution, measuring variations in echo strength betw cylindrical/spherical targets with higher/lower backscatter strengths amongst the background material
  • image recorder quality and PACs correlation

Doppler phantoms are expensive

test pattern incorporating full gray scale test display or film recorder

Ultrasound Safety

W = acoustic power output; peak power = average power/duty factor

  • calorimetric method – measuring temperature rise in medium totally absorbing transducer output
  • radiation force/pressure method – force (F) measured on suspended disc under transducer; F = 2W/v

I = intensity may be derived from W or measured directly by miniature hydrophones (tiny piezoelectric crystals) at various positions in the beam in water; peak value lies at focus in water but closer to transducer in attenuating medium

  • CW: spatial peak (ISP) at max value along central axis, spatial average (ISA) averages over cross-sectional area of beam (=W/beam area); unfocussed ISP 2-3x ISA; focussed >20x greater
  • PW: ISATA (spatial-average temporal-average), ISATP (at instant peak during pulse), ISAPA (pulse-average = ISATA x duty factor), ISPTA, ISPTP, ISPPA

Thermal Effects

absorption from beam (absorption co-efficient) causes heating, off-set by blood flow dissipation; increased with high ITA, freq, focussing, exposure time, scan mode and tissue density

ΔT of 4°C may be hazardous to foetal development, reproduced at/near bone/soft tissue interfaces with spectral pulsed Doppler

TI = thermal index = W0/Wdeg(estimated W to raise T by 1°C); defined for homogenous soft tissue (TIS), soft tissue with bone at focus (TIB) and cranial/bone at surface (TIC)


stable cavitation – below I threshold for transient cavitation submicroscopic gas bubbles grow with pressure fluctuations with resonance at a certain size producing vibrational A greater than beam

transient/collapse cavitation – minute gas bubbles grow/collapse during low/high pressure parts of cycle causing localised shock waves and thermal changes that may disrupt cells and biological macromolecules

ISPPA and peak negative pressures used; cavitation in vivo has not been proven but some diagnostic machines can exceed in vitro thresholds

MI = mechanical index = related to peak negative pressure and US f

Standing Waves

reflected waves superimposed on incident waves where gas collects at antinodes (peaks of pressure) and cells at nodes (no pressure change)

causes stasis in cells in blood vessels at the nodes damaging endothelium and leading to thrombosis

Mechanical Effects

acoustic streaming – transfer of momentum from beam (due to absorption or reflection) to liquid producing shear stressors in cells and molecules

no evidence to suggest they constitute hazard in diagnostic work


no known deleterious effects have been identified in humans as a result of diagnostic ultrasound exposure

spectral pulsed Doppler > Doppler > M-mode (long beam dwell time) > B-mode > foetal heart monitoring

scan for shortest time, lowest PRF, don’t rest transducer on skin when not scanning, use high gain/reduced power and machine design should cease output during freeze frame

ASUM (Australasian Society for Ultrasound Medicine) safety statements

to date, follow-up studies have shown no adverse health effects and obstetric US exposure, but should follow ALARA

Doppler can produce significant thermal effects, particularly near bone; nonthermal effects can result in capillary bleeding in gas-containing structures; ALARA and minimise exposure near bone/gas

CW Doppler emits low power levels and is not contra-indicated even for extended periods

thermal biological effects:

  • FDA limit is 720 mW/cm2 (usually <500mW/cm2) ISPTA estimated at tissue of interest, causing estimated max T rise of 2°C
  • exposures producing max 1.5°C rise can be used without reservation, but 4°C > normal body T (>41°C) for ≥5min should be considered potentially hazardous
  • duplex/Doppler in febrile patients may present risk to embryo/foetus; hence should minimise duration of Doppler (where ISPTA > 1W/cm2 are common) in pregnancy