If you are going to know it, then know it . . . Don't just know it.
It is one of my favorite sayings. Because the more I interact with someone online about science, the more I am understanding that most people know just enough to think they know everything but in reality, they really don't know anything about it.
I am usually the one that doesn't just barely know something, I will voraciously seek to know it inside and out. Because, I can't help myself, I am curious beyond reason.
So, here is a conversation I had with ChatGPT on ultrasounds and babies. Because, if you want to know about this topic, you might want to know as much as you can about sound and the history of sonar and ultrasound in medicine. This will take a long journey of probably hundreds of hour of video and reading. But - If I am going to know it, I am going to Know it . . . . not Just know it.
Part 1 — The physics of sound and ultrasound
1.1 What exactly is sound?
Sound is a form of mechanical energy that travels through a medium as a wave. It originates when an object vibrates, disturbing the particles around it. Those particles transfer energy to neighboring particles, allowing the disturbance to travel.
For example, when you strike a tuning fork, its prongs vibrate back and forth. As they move outward, they compress the surrounding air molecules. As they move inward, they create regions of lower pressure.
These alternating regions of compression and rarefaction travel outward as a sound wave.
Notice something important: the individual air molecules do not travel all the way from the tuning fork to your ear. They oscillate around their positions, transferring energy through the medium.
This is called a longitudinal wave because the particle movement is parallel to the direction of wave propagation.
Sound can travel through:
Gases, such as air.
Liquids, such as water.
Solids, such as steel, bone, and biological tissue.
It cannot travel through a vacuum because there are no particles available to transmit the mechanical disturbance.
This is fundamentally different from electromagnetic radiation. Light can travel through empty space; sound cannot.
Why does sound travel at different speeds?
The speed of sound depends on the physical properties of the material through which it travels.
Two major factors are:
Elasticity: How strongly a material resists compression or deformation.
Density: How much mass exists within a given volume.
Approximate speeds:
Medium | Speed of sound |
|---|---|
Air at room temperature | 343 m/s |
Water | 1,480 m/s |
Soft human tissue | 1,540 m/s |
Steel | 5,000–6,000 m/s |
These differences become important in medical imaging because the ultrasound machine has to interpret echoes returning from structures with different acoustic properties.
1.2 The five properties of sound
Frequency
Frequency is the number of complete wave cycles occurring every second. It is measured in hertz (Hz).
100 Hz = 100 cycles per second.
1,000 Hz = 1,000 cycles per second.
1 MHz = 1 million cycles per second.
Frequency determines the pitch of an audible sound.
Wavelength
Wavelength is the physical distance between corresponding points on successive waves.
Frequency and wavelength are inversely related when the speed of sound remains constant.
Where:
\(v\) = speed of sound
\(f\) = frequency
\(\lambda\) = wavelength
Amplitude
Amplitude describes the size of the pressure variation associated with the wave.
A larger amplitude generally means a larger pressure disturbance.
Intensity
Intensity describes the amount of acoustic power passing through a unit area. It is commonly expressed in watts per square centimeter (W/cm²).
This is particularly important for our safety investigation because biological interactions depend on more than frequency alone.
Propagation speed
This describes how quickly the wave travels through a particular material.
The ultrasound machine uses assumptions about sound speed to calculate the location of reflecting structures.
1.3 Infrasound, audible sound, and ultrasound
Category | Frequency |
|---|---|
Infrasound | Below 20 Hz |
Audible sound | Approximately 20 Hz–20 kHz |
Ultrasound | Above 20 kHz |
Diagnostic medical ultrasound | Commonly 1–20 MHz |
The boundaries of human hearing are approximate and vary with age and individual physiology.
Prenatal imaging uses frequencies many times higher than those needed for ordinary hearing.
For example, a 5 MHz ultrasound wave completes five million cycles every second.
But higher frequency does not automatically mean greater energy deposition or greater biological danger.
Frequency, acoustic pressure, intensity, pulse duration, exposure time, focusing, and tissue absorption all matter.
1.4 Why ultrasound is useful for imaging
The important property is not simply that ultrasound is high frequency.
It is that sound waves interact with different materials in predictable ways.
When ultrasound encounters a boundary between two tissues with different acoustic properties, some of the sound energy may be reflected back toward the transducer.
The amount reflected depends heavily on a property called acoustic impedance.
Where:
\(Z\) = acoustic impedance
\(\rho\) = density
\(c\) = speed of sound
A large difference in acoustic impedance produces a stronger reflection.
This is why ultrasound can distinguish between fluid-filled spaces, soft tissue, bone, and organs.
It also explains why ultrasound has difficulty imaging structures behind bone or air-filled organs: those boundaries reflect or scatter substantial amounts of incoming sound.
1.5 How the machine turns echoes into an image
The machine sends out a short pulse of ultrasound and then listens for returning echoes.
By measuring how long an echo takes to return, it estimates the depth of the structure that reflected it.
The division by two accounts for the round trip.
The machine also analyzes the strength of the returning echo and uses that information to determine image brightness.
This is the foundation of B-mode imaging, the familiar two-dimensional grayscale ultrasound image.
An ultrasound image is not a photograph in the ordinary sense. It is a computer-generated representation of acoustic information.
1.6 Why frequency affects image quality
Higher frequencies generally provide better spatial resolution because they have shorter wavelengths.
However, they are absorbed and scattered more strongly as they travel through tissue.
This creates a trade-off:
Higher frequency: better detail, less penetration.
Lower frequency: greater penetration, less detail.
For example, a probe designed to examine superficial structures may operate at a substantially higher frequency than one designed to image a fetus deep within the mother's abdomen.
This is one reason different transducers are manufactured for different medical applications.
Textbooks and college-level learning resources
OpenStax — College Physics 2e
Chapter 17: Sound. Particularly sections on hearing, wave properties, and ultrasound.
Section 17.7 directly covers acoustic impedance, intensity reflection, medical ultrasound, and Doppler imaging.
Read Section 17.7 — UltrasoundWHO — Manual of Diagnostic Ultrasound, Volume 1
A 430-page professional reference. Chapter 1, Basic Physics of Ultrasound, is especially relevant here.
Covers sound propagation, image formation, transducers, and the physics behind A-, B-, M-, 3D-, and 4D-mode imaging.
Read or download the WHO textbookOpenStax — University Physics, Volume 1
Useful for a deeper examination of wave mechanics and the Doppler effect.
Read the Doppler Effect chapterLecture and visual learning: OpenStax College Physics — Chapter 17 is a useful sequence of reading material and worked examples.
Part 2 — From animal echolocation to SONAR
2.1 Nature was using ultrasound long before humans understood it
Bats and toothed whales use high-frequency sound to navigate their environments.
Bats emit ultrasonic calls and interpret the returning echoes.
Depending on the species, they may use frequency-modulated calls, constant-frequency calls, or combinations of the two.
These signals allow them to determine:
Distance to an object.
Direction of an object.
Whether an object is moving.
Characteristics of potential prey.
Dolphins use a related but different biological system.
They produce short acoustic clicks and receive returning echoes through specialized structures in their lower jaw and auditory system.
These animals demonstrate that biological systems can both generate and interpret ultrasonic signals.
However, their biological systems are not identical to medical ultrasound equipment. Their sound production, signal processing, and biological purposes are different.
2.2 Early human experiments with underwater sound
The underwater environment creates a particular problem.
Light is absorbed and scattered as it travels through water, making long-distance visual observation difficult.
Sound, however, can travel considerable distances underwater.
In 1826, Swiss physicist Daniel Colladon demonstrated the transmission of sound through water in Lake Geneva.
This was an important early demonstration of underwater acoustics.
Scientists subsequently began investigating whether sound could be used to detect underwater objects.
2.3 The Titanic disaster and World War I
The sinking of the Titanic in 1912 increased interest in detecting icebergs and underwater obstacles.
Almost simultaneously, World War I created an urgent military need to detect submarines.
German U-boats were threatening Allied shipping, and conventional visual detection was inadequate.
Researchers began developing underwater acoustic detection systems.
2.4 Paul Langevin and the development of SONAR
One particularly important scientist was French physicist Paul Langevin.
During World War I, Langevin worked on developing ultrasonic submarine-detection technology.
He used piezoelectric quartz to generate and receive ultrasonic signals.
His work helped establish the pulse-echo principle that would later become central to medical imaging.
SONAR stands for Sound Navigation and Ranging.
There are two principal approaches:
Active SONAR | Passive SONAR |
|---|---|
Sends out a sound pulse | Listens for sounds |
Detects returning echoes | Detects sounds produced by other objects |
Can estimate distance | Requires additional information to estimate distance |
Similar in principle to diagnostic ultrasound | Not the normal basis of medical imaging |
Active SONAR estimates distance by measuring how long the echo takes to return.
The same fundamental principle is used in medical ultrasound.
The major difference is the purpose, equipment design, acoustic output, and environment.
Historical resources
A Historical Timeline of Diagnostic Ultrasonography — 2024.
Wiley Online Library+1British Medical Ultrasound Society — History of Ultrasound.
BMUS
The 2024 historical review is particularly useful because it connects maritime research, piezoelectricity, industrial testing, and the later development of medical ultrasound.
Part 3 — Piezoelectricity and the invention of medical ultrasound
3.1 The Curie brothers and piezoelectricity
In 1880, Pierre and Jacques Curie discovered the piezoelectric effect.
They demonstrated that certain crystals generate electrical charges when mechanical pressure is applied.
The reverse effect was subsequently demonstrated: applying an electrical field to certain materials produces mechanical deformation.
This is called the inverse piezoelectric effect.
Both effects are fundamental to conventional ultrasound transducers.
Transmission:
An electrical signal is applied to a piezoelectric material, causing it to vibrate and produce ultrasound.
Reception:
Returning sound waves cause the material to deform, producing an electrical signal that the machine processes.
The same probe can therefore function as both transmitter and receiver.
3.2 Industrial ultrasound came before widespread medical imaging
Before ultrasound became widely used in medicine, it found applications in industrial testing.
One important application was detecting defects inside metal structures.
A metal component might look perfectly normal externally but contain:
Cracks.
Voids.
Inclusions.
Internal separations.
Structural weaknesses.
Ultrasound could reveal these defects without cutting the object open.
This is called nondestructive testing.
The principle was similar to medical imaging.
Sound waves entered the material, reflected from internal boundaries, and returned to a receiver.
Researchers recognized that similar techniques might allow them to investigate structures inside the human body.
However, biological tissue presents greater complexity than a relatively uniform metal component.
3.3 The first medical experiments
During the 1940s, researchers began investigating ultrasound as a medical diagnostic tool.
One early pioneer was Karl Theodore Dussik, an Austrian neurologist.
He and his brother Friedrich explored the possibility of using ultrasound to image the brain.
Their work involved transmitting ultrasound through the head and measuring the resulting signals.
They called the technique hyperphonography.
The early images were crude, and some of the interpretations were later questioned. Nevertheless, this was an important early attempt to use ultrasound diagnostically.
Other important contributors included:
John Wild and John Reid, who investigated ultrasound in examining human tissues.
Douglas Howry and Joseph Holmes, who developed early cross-sectional imaging systems.
Inge Edler and Carl Hertz, who helped establish ultrasound applications in cardiac imaging.
Shigeo Satomura, whose work contributed to Doppler ultrasound.
3.4 Ian Donald and the development of obstetric ultrasound
If we identify one particularly influential figure in the development of obstetric ultrasound, it would be Scottish physician Ian Donald.
Donald worked at the University of Glasgow and became interested in the diagnostic possibilities of ultrasound.
He recognized that industrial ultrasonic equipment might be adapted to examine the human body.
He collaborated with engineer Tom Brown and obstetrician John MacVicar.
Their work culminated in a landmark 1958 publication:
Investigation of Abdominal Masses by Pulsed Ultrasound.
Published in The Lancet, this paper demonstrated the usefulness of ultrasound in distinguishing different abdominal structures and investigating masses.
It helped establish diagnostic ultrasound as a legitimate medical technology.
Read the original paper
The original paper is especially valuable because it lets us see what the technology could actually do at the time, rather than relying on later descriptions.
Historical video and reading
Part 4 — How ultrasound entered obstetrics
4.1 What could doctors do before prenatal ultrasound?
For much of the nineteenth and early twentieth centuries, doctors had limited ways of assessing the developing fetus.
Pregnancy was evaluated indirectly through:
Measurement of the mother's abdomen.
Listening for fetal heart sounds.
Feeling fetal movement.
Physical examination to determine fetal position.
Estimation of gestational age.
X-ray examinations in certain historical circumstances.
X-rays could reveal the position of a fetal skeleton, but they could not provide the soft-tissue imaging that ultrasound eventually made possible.
Ultrasound changed this because it provided a way to examine the developing fetus without ionizing radiation.
That was a major advantage.
But we should not confuse non-ionizing radiation with biological inactivity. Ultrasound is mechanical energy and has its own biological interactions.
4.2 Early obstetric applications
The initial clinical applications included distinguishing pregnancy from certain abdominal masses and examining the uterus.
As the technology improved, doctors could assess:
Fetal number.
Fetal position.
Fetal growth.
Gestational age.
Placental location.
Certain structural abnormalities.
Fetal cardiac activity.
4.3 The evolution of the equipment
Period | Development |
|---|---|
1940s | Early experimental medical ultrasound |
1950s | Clinical abdominal and obstetric applications |
1960s | Commercial scanners become available |
1970s | Improved grayscale imaging |
1980s | Real-time imaging becomes increasingly practical |
1990s | More advanced electronics and Doppler capabilities |
2000s | Wider availability of 3D and 4D imaging |
2010s–2020s | High-resolution imaging, portable systems, and advanced image processing |
The transition from still images to real-time imaging was particularly important.
Doctors could observe fetal movement, cardiac activity, and blood flow rather than relying only on static images.
4.4 The different types of prenatal ultrasound
This is important because the term ultrasound describes several different imaging techniques.
They do not all operate in exactly the same way.
B-mode — Brightness mode
Produces the familiar two-dimensional grayscale image.
Echo strength determines pixel brightness. It is used to examine fetal anatomy, growth, and position.
M-mode — Motion mode
Records movement along a particular line through tissue.
It can be used to document cardiac activity and measure heart rate.
Color Doppler
Uses frequency changes in returning echoes to display information about blood movement.
Red and blue indicate direction relative to the transducer, not inherently oxygenated versus deoxygenated blood.
Spectral Doppler
Displays blood-flow velocity over time as a waveform.
It can evaluate circulation in the umbilical artery and other vessels.
3D and 4D ultrasound
Three-dimensional imaging reconstructs a volume of anatomical information.
Four-dimensional imaging adds time, allowing the reconstructed volume to be displayed as moving images.
Why the distinction matters
Different imaging modes can have different acoustic outputs.
B-mode and M-mode generally have lower time-averaged acoustic output than many Doppler applications.
Pulsed Doppler can concentrate acoustic energy in a relatively small region to measure blood flow.
This is why professional organizations have specific recommendations about Doppler use, particularly during early pregnancy.
We will examine the actual research and recommendations in the safety chapters.
Technical textbook: The WHO's Manual of Diagnostic Ultrasound, Volume 1, covers these imaging modes in its first chapter.
WHO — Manual of Diagnostic Ultrasound, Volume 1.
Part 5 — How ultrasound interacts with biological tissue
This is where our investigation moves from engineering into biology.
We know how ultrasound is generated and how echoes produce an image.
But what happens to the energy after it enters the body?
Does it simply bounce around until it returns to the transducer?
No.
Some is reflected, some is scattered, some passes deeper into tissue, and some is absorbed.
The absorbed energy can produce biological effects.
5.1 Reflection, transmission, scattering, and absorption
Reflection: A portion of the sound wave returns toward the transducer when it encounters a boundary between tissues with different acoustic impedances.
Transmission: Some acoustic energy continues through the boundary into the next tissue.
Scattering: Sound is redirected in different directions when it encounters small structures or irregularities.
Absorption: Some acoustic energy is converted into other forms of energy, primarily heat.
Absorption is particularly important when investigating biological effects.
5.2 Why ultrasound produces heat
When an acoustic wave travels through tissue, pressure variations cause particles and structures within the tissue to move.
Some mechanical energy is dissipated and converted into thermal energy.
The amount of heating depends on:
Acoustic intensity.
Frequency.
Tissue composition.
Absorption characteristics.
Duration of exposure.
Beam concentration.
Blood circulation and heat removal.
Bone absorbs ultrasound differently from many soft tissues.
This is particularly relevant during fetal imaging because the developing skeleton becomes an increasingly important acoustic target as pregnancy progresses.
What does heating mean biologically?
Temperature influences biological processes.
As temperature increases:
Chemical reaction rates can change.
Cellular membranes can become more fluid.
Protein structures can be affected.
Cellular stress responses can be activated.
At sufficiently high temperatures and exposure durations, cells can be injured or killed.
However, we need to distinguish between:
A measurable temperature increase.
A temperature increase that changes biological processes.
A temperature increase sufficient to cause tissue injury.
These are not interchangeable.
A small temperature increase does not automatically mean injury has occurred.
Researchers therefore consider thermal dose, which takes both temperature and duration into account.
5.3 Mechanical effects
The second major category involves mechanical interactions.
Ultrasound produces alternating regions of positive and negative acoustic pressure.
These pressure changes can cause physical movement and deformation within tissues and fluids.
Important mechanisms include:
Acoustic radiation force
The sound wave transfers momentum to the material through which it travels, producing a small force on tissue.
Acoustic streaming
The wave can generate fluid movement.
Cavitation
This involves the behavior of microscopic gas bodies exposed to alternating acoustic pressure.
There are two major forms:
Stable cavitation involves oscillation of a gas bubble.
Inertial cavitation involves more violent expansion and collapse.
Bubble collapse can generate localized mechanical forces and, under some circumstances, very high local temperatures.
This sounds alarming, but context is essential.
Cavitation is strongly influenced by the presence of gas nuclei, acoustic pressure, frequency, and other conditions.
Routine prenatal ultrasound does not involve injecting contrast microbubbles.
We should not assume that cavitation observed in specialized laboratory experiments automatically occurs during ordinary fetal imaging.
5.4 Can ultrasound affect cells and organs?
Researchers have investigated ultrasound exposure in:
Cultured cells.
Animal tissues.
Embryonic and fetal animal models.
Blood vessels.
Lung tissue.
Nervous tissue.
One established experimental bioeffect is pulmonary capillary hemorrhage in certain laboratory animal models.
The AIUM describes this phenomenon and notes that the biological effects depend on the tissue, acoustic output, exposure conditions, and presence of gas bodies.
The lungs are gas-containing organs, and gas-liquid interfaces behave differently under ultrasound exposure.
A developing fetus in the uterus has a different biological environment.
We cannot simply take an experiment involving an animal's lung and conclude that a human fetus experiences the same effect.
This is why different tissue models and exposure conditions matter.
5.5 Diagnostic versus therapeutic ultrasound
Diagnostic ultrasound
Designed to obtain information about the body.
Generally uses controlled acoustic output and pulses to generate images.
Therapeutic ultrasound
Designed to produce a biological or physical effect.
May deliberately generate heating, mechanical disruption, or other tissue effects.
Examples of therapeutic ultrasound include:
High-intensity focused ultrasound (HIFU).
Some physical therapy applications.
Ultrasound-assisted surgical procedures.
Experimental ultrasound neuromodulation.
Research demonstrating tissue destruction using therapeutic ultrasound does not demonstrate that diagnostic prenatal ultrasound causes similar damage.
Likewise, evidence that diagnostic ultrasound has not produced a particular adverse outcome does not establish that all ultrasound exposures are harmless regardless of intensity or duration.
Research and textbook resources for this chapter
WHO — Manual of Diagnostic Ultrasound, particularly Chapter 1.
World Health Organization+1AIUM — Statement on Biological Effects of Ultrasound In Vivo.
www.aium.org+1Ultrasound Biophysics Mechanisms — NCBI/PMC.
PMC
Part 6 — The development of ultrasound safety standards
Now we reach the first major safety investigation.
How did scientists and regulators decide what levels of ultrasound output were acceptable for medical use?
And how did those standards change as the technology improved?
6.1 The early assumption of safety
One interesting aspect of ultrasound history is that the technology entered clinical medicine before modern methods of evaluating every possible biological effect were established.
It was attractive because it did not use ionizing radiation.
X-rays had already demonstrated that radiation could damage biological tissue and cause developmental abnormalities at sufficient exposures.
Ultrasound offered a way to visualize internal structures without that particular hazard.
But the absence of ionizing radiation does not mean the absence of biological effects.
Ultrasound is mechanical energy, and researchers gradually began investigating whether it could produce:
Heating.
Mechanical disturbances.
Changes in cellular function.
Effects on blood vessels.
Changes in embryonic and fetal development.
An important historical point is that safety standards did not emerge from one definitive experiment. They developed through regulatory decisions, engineering measurements, laboratory studies, and clinical experience.
6.2 The 1976 Medical Device Amendments
In 1976, the United States passed the Medical Device Amendments to the Federal Food, Drug, and Cosmetic Act.
This established a more formal regulatory framework for medical devices.
Diagnostic ultrasound equipment was included in this framework.
The FDA subsequently developed procedures for evaluating and regulating ultrasound devices, including acoustic output.
However, the original approach was not based on a comprehensive demonstration that every permitted output was biologically harmless.
As Douglas Miller explains in his 2008 review, the initial safety framework relied substantially on the equipment already in clinical use when the legislation was enacted.
This is an important distinction: regulatory acceptance and experimental proof of zero biological risk are not the same thing.
6.3 The 1985 and 1987 acoustic output limits
In 1985, the FDA established application-specific guidance for diagnostic ultrasound equipment.
Different applications had different acoustic output limits.
Fetal imaging was assigned a relatively low limit.
In 1987, the permitted values for fetal imaging were increased.
According to Miller's historical review, the fetal imaging spatial-peak temporal-average intensity limit changed from:
46 mW/cm² in 1985.
94 mW/cm² in 1987.
These are equipment-output limits, not measurements of the energy absorbed by a fetus.
That distinction will become important when we discuss how sound travels through the mother's body.
6.4 The 1992 change: A particularly important development
In 1992, a new system called the Output Display Standard (ODS) was adopted.
This introduced the display of two important indices on ultrasound machines:
Thermal Index (TI).
Mechanical Index (MI).
The purpose was to provide operators with real-time information about potential biological effects.
The new framework permitted higher maximum acoustic outputs, with output indices displayed to help the operator make informed decisions.
For fetal imaging, the permitted derated spatial-peak temporal-average intensity under the newer framework could reach 720 mW/cm², compared with the earlier 94 mW/cm² Track 1 limit.
That is approximately a 7.7-fold difference in the permitted maximum.
Historical comparison of permitted output
Fetal imaging, spatial-peak temporal-average intensity (derated)
94 → 720
1987 Track 1 limit
1992 Track 3 framework
mW/cm². These are regulatory maximum values, not typical fetal exposures or measured absorbed doses.
This is one of the most important developments in our investigation.
But there are qualifications.
The increase in permitted output does not mean every modern ultrasound machine delivers 7.7 times the energy of an older machine.
The newer framework permitted greater output under specified conditions and required relevant output information to be displayed.
Actual exposure varies with:
Machine model.
Transducer.
Imaging mode.
Output settings.
Pulse repetition frequency.
Focusing.
Scanning time.
Tissue properties.
The distinction between permitted maximum output and actual clinical exposure is essential.
The original safety review
Douglas L. Miller published:
Safety Assurance in Obstetrical Ultrasound.
Published in Seminars in Ultrasound, CT and MRI in 2008, this paper examines the development of safety standards and the evidence supporting them.
One particularly important observation is that systematic fetal safety research using modern equipment had been limited.
The paper discusses the reliance on experimental bioeffects research, theoretical modeling, and the training and judgment of professional operators.
This is not proof that modern ultrasound is unsafe. It is an important discussion of the evidence base and its limitations.
Read the full paper:
6.5 Understanding the Thermal Index
The Thermal Index is an indicator related to the potential for temperature elevation during ultrasound exposure.
It is not a direct thermometer.
A displayed TI of 1 does not mean that the tissue has necessarily increased by exactly 1°C.
It is an estimate based on acoustic output and assumptions about tissue absorption and heating.
There are three principal versions:
Index | Meaning |
|---|---|
TIS | Thermal Index for Soft Tissue |
TIB | Thermal Index for Bone |
TIC | Thermal Index for Cranial Bone |
Why are there different indices?
Because different tissues absorb ultrasound differently.
Bone can absorb substantially more acoustic energy than many soft tissues.
As fetal development progresses and more fetal bone becomes present, the possibility of localized heating near bone becomes particularly relevant.
6.6 Understanding the Mechanical Index
The Mechanical Index relates to the potential for mechanical bioeffects.
It is calculated using the peak rarefactional pressure and the square root of the ultrasound frequency.
Where:
\(P_{r.3}\) = derated peak rarefactional pressure in MPa.
\(f_c\) = center frequency in MHz.
Like the Thermal Index, the Mechanical Index is an indicator, not a direct measurement of tissue injury.
It helps the operator understand the potential for mechanical effects under particular operating conditions.
6.7 The ALARA principle
ALARA stands for:
As Low As Reasonably Achievable.
The principle is to obtain the necessary diagnostic information using the lowest practical acoustic output and shortest practical exposure.
This does not mean turning the machine down so far that a useful examination becomes impossible.
It means avoiding unnecessary exposure and using the equipment appropriately.
The AIUM recommends that obstetric scans begin at a displayed TI of 0.7 and that higher outputs be used only when needed to obtain adequate images. It also advises against routine use of pulsed Doppler without a clinical indication.
Regulatory and technical reading
FDA — Marketing Clearance of Diagnostic Ultrasound Systems and Transducers.
U.S. Food and Drug AdministrationHow to Interpret the Ultrasound Output Display Standard.
Wiley Online Library
Part 7 — Animal experiments and fetal development
This is where I want us to be particularly careful about separating experimental findings from claims about human pregnancy.
Animal research is essential because researchers can control exposure conditions and examine tissue directly.
They can expose animals at particular developmental stages, measure temperature, examine organs, and investigate cellular changes.
But an effect observed in an animal does not automatically establish that the same effect occurs in human pregnancies.
7.1 What are researchers looking for?
Animal ultrasound experiments have investigated several possible effects.
Thermal effects
Temperature increases in embryonic and fetal tissues.
Localized heating near bone.
Changes in cellular development.
Effects on organ formation.
Mechanical effects
Changes in cell membranes.
Effects on blood vessels.
Mechanical stress in tissues.
Cavitation-related effects.
Developmental effects
Changes in fetal growth.
Changes in neurological development.
Changes in behavior after birth.
Effects on organ development.
7.2 Why developmental timing matters
Embryonic and fetal development are not uniform processes.
Different structures develop at different times.
Early pregnancy involves rapid cell division and organ formation.
Later pregnancy involves growth, maturation, and functional development.
Therefore, an exposure during early organ development might not have the same consequences as an exposure later in pregnancy.
This is why animal experiments need to specify:
Species.
Developmental stage.
Exposure frequency.
Acoustic pressure.
Intensity.
Duration.
Imaging mode.
Temperature changes.
Without these details, it is difficult to interpret a study.
7.3 What about neurological development?
One area of research has examined whether ultrasound exposure could affect the developing nervous system.
Some animal experiments have investigated neuronal migration and brain development.
A frequently discussed topic is whether prolonged ultrasound exposure during development could alter the distribution of neurons in the developing brain.
However, findings from laboratory models must be interpreted carefully.
Experimental exposures may differ substantially from clinical prenatal examinations in duration, acoustic output, beam characteristics, and tissue conditions.
We need to examine the actual experimental protocols before deciding how relevant a particular finding is.
7.4 The AIUM review of experimental biological effects
The American Institute of Ultrasound in Medicine published a detailed statement in 2021 reviewing biological effects observed in laboratory mammals.
It discusses how the likelihood of effects depends on:
Acoustic intensity.
Output power.
Beam focusing.
Frequency.
Peak rarefactional pressure.
Pulse repetition frequency.
Pulse duration.
Exposure duration.
Tissue type.
Presence of gas bodies.
Tissue perfusion and heat removal.
The statement is especially valuable because it does not simply declare ultrasound safe or unsafe.
It describes the conditions under which biological effects have been observed.
AIUM — Statement on Biological Effects of Ultrasound In Vivo.
7.5 An important example: Pulmonary capillary hemorrhage
Research has shown that diagnostic-range ultrasound can produce pulmonary capillary hemorrhage in certain animal models.
This is a real experimental finding.
The AIUM reports that the likelihood depends on exposure conditions and the presence of gas-containing tissue structures.
It also explains that the translation of these findings to human clinical conditions is difficult.
The human health implications of this effect in certain examinations remain uncertain.
This is an excellent example of why we cannot simply classify every ultrasound bioeffect as either irrelevant or proof of harm to a fetus.
Research resources
For this chapter, I would particularly like to build a separate annotated collection of original animal experiments. The AIUM statement is a useful entry point, but we should examine the cited experiments individually rather than treating a professional organization's summary as a substitute for the primary research.
Part 8 — Human pregnancy studies: What has actually been demonstrated?
This is the most important part of the safety investigation.
Animal studies can identify possible mechanisms of injury.
But the question parents ultimately care about is whether prenatal ultrasound exposure has been shown to cause harm in human pregnancies.
We need to examine the human evidence.
8.1 Observational studies
Observational studies examine pregnancies that have already experienced ultrasound exposure.
Researchers may compare:
Birth weight.
Gestational age.
Pregnancy complications.
Congenital abnormalities.
Childhood development.
Neurological outcomes.
These studies are useful but have limitations.
For example, women who receive more ultrasound examinations may have pregnancies considered higher risk.
If their children experience more complications, it does not necessarily mean ultrasound caused them.
The underlying medical condition may explain the difference.
This is called confounding.
Researchers also need to consider whether exposure has been measured accurately.
A medical record may show that a woman received three ultrasound examinations, but it may not contain sufficient information about:
Acoustic output.
Imaging mode.
Time spent imaging specific structures.
Thermal Index.
Mechanical Index.
Actual energy absorbed by fetal tissue.
This makes some historical studies difficult to interpret.
8.2 The Newnham 1993 randomized controlled trial
One particularly important study was published in The Lancet in 1993.
Title: Effects of frequent ultrasound during pregnancy: a randomised controlled trial.
Researchers enrolled 2,834 women with singleton pregnancies.
They randomly assigned participants to two groups.
Group | Examination schedule |
|---|---|
Intensive group | Ultrasound imaging and continuous-wave Doppler at 18, 24, 28, 34, and 38 weeks |
Regular group | One ultrasound examination at 18 weeks |
This was important because it directly investigated repeated prenatal ultrasound exposure rather than merely comparing women who had received scans with women who had not.
The study found a higher proportion of babies classified as small for gestational age in the intensive group.
The reported relative risk for birth weight below the 10th centile was 1.35, with a 95% confidence interval of 1.09–1.67.
This finding deserves attention.
However, it does not establish that ultrasound caused impaired fetal growth.
The study involved a particular examination schedule, older equipment, and continuous-wave Doppler.
It cannot automatically answer every question about modern equipment or every imaging mode.
Read the original research:
8.3 The eight-year follow-up
In 2004, the researchers published a follow-up examining the children from the original trial.
Title: Effects of repeated prenatal ultrasound examinations on childhood outcome up to 8 years of age.
They evaluated:
Physical growth.
Neurological development.
Speech.
Language.
Behavior.
Childhood development.
The researchers did not identify significant adverse developmental outcomes attributable to the intensive ultrasound schedule.
This is valuable evidence because it extends beyond birth outcomes.
However, follow-up studies also have limitations, including loss of participants over time.
The study followed a particular historical exposure protocol, not every possible configuration of modern ultrasound equipment.
Read the original research:
- Newnham et al. (2004) — PubMed.
8.4 Routine versus selective ultrasound: The Cochrane evidence
Another important source is the Cochrane review comparing routine early pregnancy ultrasound with selective ultrasound.
It examined randomized trials involving tens of thousands of women.
The review found that routine ultrasound improved detection of multiple pregnancies and major fetal abnormalities.
It did not demonstrate a reduction in perinatal mortality.
This is important because it helps separate the benefits of screening from claims about improved survival.
It is also important to understand what the review was designed to measure.
It was primarily concerned with clinical outcomes of routine versus selective scanning, not a comprehensive investigation of every possible biological effect of ultrasound.
.
8.5 What about autism, dyslexia, and handedness?
There have been observational studies investigating possible associations between prenatal ultrasound exposure and various childhood outcomes.
Some have examined:
Handedness.
Speech development.
Dyslexia.
Autism-related outcomes.
Behavioral differences.
These studies have not established that prenatal diagnostic ultrasound causes these conditions.
There are several difficulties in interpreting such research:
Older equipment may have different output characteristics.
Exposure records may be incomplete.
Women receiving more scans may have higher-risk pregnancies.
Outcomes may be influenced by genetic and environmental factors.
Associations do not establish causation.
I would not dismiss these studies simply because their findings are inconvenient, but I also would not treat an observed association as evidence of injury.
The proper approach is to look at study design, sample size, exposure measurement, confounders, replication, and the broader literature.
8.6 What do the major medical organizations conclude?
American College of Obstetricians and Gynecologists (ACOG)
ACOG identifies ultrasound as an imaging technique of choice during pregnancy when it is expected to answer a relevant clinical question or provide medical benefit.
It recommends prudent use and the ALARA principle.
Its guidance also recognizes that B-mode imaging generally has lower potential for heating than color and spectral Doppler.
ACOG — Guidelines for Diagnostic Imaging During Pregnancy and Lactation.
American Institute of Ultrasound in Medicine (AIUM)
AIUM recognizes biological effects in experimental systems but states that independently confirmed adverse effects from current diagnostic instruments have not been established in human patients in the absence of contrast agents.
It recommends prudent clinical use.
AIUM — Prudent Use and Safety of Diagnostic Ultrasound in Pregnancy.
International Society of Ultrasound in Obstetrics and Gynecology (ISUOG)
ISUOG has issued specific recommendations for Doppler use during early pregnancy.
It recognizes that Doppler modalities can have greater energy output and therefore require additional attention to exposure time and thermal indices.
ISUOG — Safe Use of Doppler in the First 13+6 Weeks of Pregnancy.
These organizations are useful sources of professional guidance, but we should continue to give priority to original research when evaluating specific claims.
Part 9 — Modern ultrasound machines and the differences in exposure
This chapter brings together the physics, technology, and safety research.
One of the things I want to avoid is treating every ultrasound examination as though it delivers the same exposure.
It doesn't.
The imaging mode, transducer, output settings, pulse characteristics, and examination duration all influence acoustic exposure.
9.1 B-mode versus Doppler
B-mode primarily uses short pulses to create anatomical images.
Doppler techniques analyze frequency shifts in returning echoes to assess movement, particularly blood flow.
The distinction is important because Doppler applications can have higher time-averaged acoustic output.
Imaging technique | Primary purpose | Safety consideration |
|---|---|---|
B-mode | Anatomical imaging | Generally lower heating potential |
M-mode | Movement and cardiac activity | Exposure depends on operating conditions |
Color Doppler | Direction and distribution of blood flow | Can have higher acoustic output |
Power Doppler | Detecting blood-flow signals | Output and exposure duration matter |
Spectral Doppler | Measuring blood-flow velocity | Requires particular attention to exposure |
3D/4D | Reconstructing anatomical volumes and movement | Depends on acquisition mode and settings |
9.2 Why Doppler deserves particular attention
Doppler ultrasound repeatedly samples returning acoustic signals to detect movement.
Depending on the system and operating mode, this can result in greater time-averaged exposure than conventional B-mode imaging.
This is particularly relevant during early pregnancy.
The ISUOG distinguishes between the embryonic and fetal periods.
Its 2021 guidance explains that during the embryonic period, which it treats as extending through 10 weeks and 6 days by last-menstrual-period dating, Doppler should not be used routinely.
If there is a clinical reason to use it, exposure time should be minimized.
From 11 weeks through 13 weeks and 6 days, Doppler may be used for specified clinical indications, with a displayed TI of 1.0 or lower and exposure kept as short as possible, usually no longer than 5–10 minutes.
These are professional recommendations, not evidence that every exposure above those values causes harm.
Read the complete ISUOG safety statement.
9.3 What about 3D and 4D ultrasound?
This is particularly interesting because many parents associate 3D and 4D imaging with keepsake photographs and videos.
Three-dimensional ultrasound reconstructs a volume of anatomical information.
Four-dimensional ultrasound displays that volume over time.
But the appearance of a 3D image does not tell us how much energy was delivered.
We need to investigate:
How the volume was acquired.
Whether Doppler was used.
How long the probe remained in a particular position.
The output settings.
The thermal index.
The actual examination duration.
The technology should be assessed by its acoustic exposure, not simply by whether the image is two-dimensional or three-dimensional.
9.4 Medical imaging versus recreational scanning
There is a distinction between an examination intended to answer a medical question and one intended primarily to obtain photographs or videos.
Medical examinations are designed around diagnostic objectives.
Recreational examinations may involve additional scanning time without providing corresponding medical benefit.
That does not establish that a keepsake scan causes injury.
It does mean that additional exposure should not be assumed to be beneficial or necessary.
The FDA advises against nonmedical ultrasound imaging and fetal heartbeat monitoring for keepsake purposes.
Professional guidance:
9.5 What about fetal heartbeat monitors at home?
These devices deserve a separate mention.
Some handheld fetal Doppler devices are marketed to pregnant consumers.
They use ultrasound to detect fetal cardiac movement and produce an audible representation of the signal.
There are two important issues:
The consumer may not understand the acoustic exposure or how to use the device appropriately.
The sound heard through the device is not a reliable substitute for professional assessment of fetal well-being.
A heartbeat detected by a consumer device does not establish that the fetus is healthy.
Conversely, failure to find a heartbeat can cause unnecessary distress.
The FDA advises against using fetal ultrasound heartbeat monitors without medical supervision.
This is another reason to distinguish medical diagnostic ultrasound from recreational or unsupervised use.
Part 10 — What does the evidence actually tell us?
Now we can bring the investigation together.
I would divide our conclusions into four categories.
10.1 What is well established?
Ultrasound is mechanical energy, not ionizing radiation.
Diagnostic ultrasound can produce thermal and mechanical biological effects.
The amount of energy delivered depends on frequency, intensity, pressure, duration, beam characteristics, and imaging mode.
Different ultrasound modes can have substantially different acoustic output.
Doppler applications generally deserve greater attention to exposure than routine B-mode imaging.
Animal studies have demonstrated certain biological effects under particular experimental conditions.
Decades of clinical use and human research have not established convincing evidence that routine diagnostic prenatal ultrasound causes fetal developmental injury.
10.2 What is reassuring?
The randomized trials examining repeated prenatal ultrasound have not demonstrated convincing evidence of long-term childhood developmental harm.
The Newnham 2004 follow-up is particularly valuable because it examined children up to eight years of age.
Professional organizations consider medically indicated diagnostic ultrasound an acceptable method of imaging during pregnancy.
The evidence supporting routine clinical use is substantial.
10.3 What deserves further research?
I think these are the questions we should not simply gloss over:
How much modern equipment output actually reaches fetal tissue?
What are the temperature changes in different fetal structures under different scanning conditions?
How do repeated modern Doppler examinations compare with older exposure protocols?
What are the long-term outcomes of repeated 3D/4D examinations?
Are there meaningful differences between early embryonic and later fetal exposure?
How well do historical human studies represent today's ultrasound machines?
What is the clinical significance of biological effects identified in animal experiments?
One important limitation is that the absence of a demonstrated human adverse effect does not establish that every possible exposure has been studied.
That is why the ALARA principle remains important.
10.4 My overall assessment
The current evidence supports the use of diagnostic prenatal ultrasound when there is a legitimate medical indication and the examination is performed appropriately.
I would not characterize routine prenatal ultrasound as a demonstrated danger to babies.
At the same time, I would not describe it as biologically inert or suggest that unlimited exposure is harmless.
The distinction is between a technology that has demonstrated biological interactions and a technology that has demonstrated clinically significant harm under ordinary prenatal diagnostic conditions.
Those are two very different scientific conclusions.
Complete research and learning library
Here is how I would organize the resources so that we can keep adding to this project.
Subject | Primary resource | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Basic physics | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Advanced wave physics | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Ultrasound textbook | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Historical development | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
First medical applications | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Biological effects | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Regulatory history | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Animal research | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Repeated prenatal scans | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Childhood outcomes | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Clinical guidance | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Early Doppler | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Prudent use 1. Acoustic energy transmission from mother to fetusThis is where the safety question becomes much more interesting, because the acoustic output reported by the ultrasound machine is not the same thing as the acoustic energy reaching the fetus. The machine creates an acoustic field at the transducer. That field then changes as it travels through tissue. Some energy is reflected at boundaries, some is scattered, and some is absorbed and converted into heat. The amount that continues toward the fetus depends on the frequency, the tissue layers, their thicknesses, the beam's angle and focus, and the ultrasound mode and settings. We can follow the energy path, but we need to keep two questions separate:
Those are related, but they are not interchangeable. A tissue can transmit much of the beam while absorbing a smaller fraction of it; another tissue can absorb more energy and therefore receive more local heating. A. First, what exactly are we measuring?For this exercise, the most useful quantities are:
A useful distinction: intensity is not the same as total energy delivered. To estimate total energy, we also need to know how long the beam is present, how large the insonated area is, and whether the beam is continuously on or pulsed. For example, a machine may display a relatively high instantaneous or spatial-peak output, but the time-averaged intensity can be much lower because the beam is pulsed and the transducer scans across an image. That difference matters enormously when we try to estimate fetal exposure. B. The path from the transducer to the fetusA typical transabdominal scan sends sound through approximately this sequence:
The actual sequence varies with fetal position, placental position, maternal anatomy, and where the sonographer places the probe. The beam may also travel through different combinations of tissue rather than one neat stack of layers. 1. The gel and skinThe gel is not just there to help the probe slide. It solves a significant acoustic problem. Air and human tissue have very different acoustic impedances. At an air–skin boundary, nearly all the incident ultrasound energy would be reflected. Gel removes the microscopic air pockets between the transducer and skin and provides a much better acoustic connection. Once the wave enters the skin, it encounters a series of tissue boundaries. The impedance differences between ordinary soft tissues are relatively small, so reflection at many of these boundaries is modest. The more important loss over a long path is often attenuation within the tissues themselves. 2. Subcutaneous fatFat generally attenuates ultrasound less than muscle, although its properties vary with tissue composition and frequency. It is tempting to assume that a thicker fat layer simply blocks more ultrasound. It does reduce the beam, but the relationship depends on the attenuation coefficient and frequency—not thickness alone. 3. Abdominal muscleMuscle can attenuate ultrasound more strongly than fat. Its structure is also directional: the acoustic properties can differ depending on whether the beam travels parallel or perpendicular to muscle fibers. This is one reason a single generic attenuation number cannot accurately describe every pregnancy. 4. The uterusThe uterine wall adds another tissue layer. An especially useful source here is an actual human in-vivo study of the acoustic path during gynecologic ultrasound. Researchers estimated tissue-specific attenuation coefficients using measurements from 162 independent paths. Their reported averages were:
These values are particularly relevant because they were measured in living humans along reproductive-ultrasound paths, rather than simply borrowed from a generic tissue table. The researchers also compared these results with the standard FDA homogeneous-tissue model. PubMed 5. Amniotic fluidHere is where the path gets interesting. Amniotic fluid behaves acoustically much more like water than like muscle or skin. Its attenuation is very low. A WHO reference table gives an approximate attenuation value for water at 1 MHz of 0.0022 dB/cm. Amniotic fluid is listed as having a similarly low attenuation value. Iris That means the amniotic fluid does not remove much energy from the beam as it travels through it. It is not an acoustic shield in the same sense that a strongly attenuating tissue would be. This is an important distinction: the fetus is not protected simply because the beam has passed through the mother's tissues and then enters fluid. The fluid can transmit sound quite efficiently. 6. Fetal tissuesOnce the beam reaches the fetus, it encounters skin, fat, muscle, organs, blood, and potentially bone. The beam continues to lose energy as it travels through fetal tissues. Some energy is reflected back toward the transducer, some is scattered in different directions, and some is absorbed. Bone deserves special attention. Its acoustic impedance differs substantially from soft tissue, so a soft-tissue-to-bone boundary can reflect a considerable fraction of the incident intensity. Bone also absorbs ultrasound more strongly than most soft tissues. That does not mean every fetal bone receives a large exposure. The result depends on whether the beam actually intersects the bone, the angle of incidence, the beam's focal location, and the operating mode. C. The mathematics: how much intensity is lost?The basic attenuation equation is: \[
I(x)=I_0\,10^{-A/10}
\] Where:
For a simple tissue layer: \[
A=\alpha f d
\] Where:
For multiple layers, we add the attenuation in each layer: \[
A_{\text{total}}=\sum_i \alpha_i f d_i
\] This gives us a useful first approximation of how much intensity remains in the forward-traveling beam. The equation also shows why frequency matters so much. If the attenuation coefficient is approximately proportional to frequency, doubling the frequency doubles the attenuation in decibels for the same tissue path. A worked exampleLet's construct a hypothetical transabdominal scan. These are illustrative layer thicknesses, not measurements from a particular pregnancy. Assume a 3 MHz beam and the following path:
We calculate each layer's attenuation: Skin and subcutaneous tissue \[
2.3\times3\times2.0=13.8\text{ dB}
\] Muscle \[
3.1\times3\times1.0=9.3\text{ dB}
\] Myometrium \[
0.6\times3\times0.5=0.9\text{ dB}
\] Amniotic fluid \[
0.0022\times3\times3.0=0.0198\text{ dB}
\] Total: \[
A_{\text{total}}=24.02\text{ dB}
\] The intensity remaining at the fetal surface, under these assumptions, is: \[
\frac{I}{I_0}=10^{-24.02/10}
\] \[
\boxed{I/I_0\approx0.00396}
\] Illustrative result Approximately 0.4% of the starting intensity remains in the forward beam at the fetal surface in this particular layered model. That is a large reduction. But we need to be very careful about what it means. It does not mean the fetus receives exactly 0.4% of the machine's displayed output. The calculation assumes a particular frequency, path, and set of attenuation coefficients. It also does not yet account for the transducer's beam profile, focal characteristics, reflections, or the distinction between peak and time-averaged intensity. And importantly, the attenuation coefficients from the in-vivo study are averages, not universal constants. Individual paths can differ substantially. What happens if we use a different attenuation model?This is a particularly important issue. There isn't just one accepted way to estimate the in-body acoustic field. The FDA uses a standardized attenuation assumption of 0.3 dB/cm/MHz when reporting certain derated acoustic-output quantities. This is a regulatory comparison model, not a measurement of the actual tissues in an individual pregnant patient. U.S. Food and Drug Administration+1 For our 6.5 cm path from the transducer to the fetal surface, at 3 MHz: \[
A=0.3\times3\times6.5=5.85\text{ dB}
\] That leaves: \[
10^{-5.85/10}=0.260
\] Or about 26% of the starting intensity. Now compare three approaches:
These estimates are dramatically different. The lesson is not that one of these percentages is automatically the correct fetal exposure. The lesson is that a generic derating factor and a patient-specific tissue-path calculation are answering somewhat different questions. The in-vivo research is especially useful because it demonstrates that tissue-specific attenuation can vary substantially, and that the choice of model has a major effect on the estimated exposure. PubMed+1 This is one of the limitations we should keep in mind whenever someone quotes a machine's displayed output as though it were the intensity at the fetus. D. Reflection: how much energy bounces back?Attenuation is only part of the story. At every boundary between two materials, some of the acoustic wave is reflected. The amount of reflection depends on the difference in acoustic impedance. Acoustic impedance is: \[
Z=\rho c
\] Where:
For a wave striking a flat boundary at normal incidence, the intensity reflection coefficient is: \[
R=\left(\frac{Z_2-Z_1}{Z_2+Z_1}\right)^2
\] Let's use approximate impedances to see what happens.
MRayl is one million Rayl. Fat-to-muscle boundary\[
R=\left(\frac{1.7-1.4}{1.7+1.4}\right)^2
\] \[
R\approx0.0094
\] So approximately 0.94% of the intensity is reflected at this idealized boundary. Most continues through. Soft tissue-to-amniotic-fluid boundaryUsing 1.6 and 1.5 MRayl: \[
R=\left(\frac{1.5-1.6}{1.5+1.6}\right)^2
\] \[
R\approx0.00104
\] Approximately 0.1% is reflected. Again, most continues through. Soft tissue-to-bone boundaryUsing 1.6 and 6.3 MRayl: \[
R=\left(\frac{6.3-1.6}{6.3+1.6}\right)^2
\] \[
R\approx0.354
\] Approximately 35% is reflected in this simplified example. This is why bone can create strong echoes and acoustic shadowing. The numbers are idealized calculations. Real boundaries are not perfectly flat, and the angle of the beam, tissue structure, and surface irregularities can change the result. The University of Michigan's biomedical physics teaching material provides a useful discussion of impedance and reflection at tissue boundaries. socr.umich.edu+1 One thing I want to emphasize: reflection does not mean the reflected energy disappears. It travels in another direction. Some of it returns to the transducer and becomes the signal used to construct the image. Some may be redirected elsewhere. E. Absorption: where the energy actually goesThis is the part most directly connected to heating. When ultrasound is absorbed, the mechanical energy of the sound wave is transferred into the tissue. Much of that energy ultimately becomes heat. But attenuation and absorption are not identical:
A review of quantitative ultrasound describes attenuation as a combination of absorption and scattering, with additional effects from reflection, refraction, and diffraction. PMC+1 This distinction matters because a 10 dB reduction in the forward beam does not necessarily mean that all the missing energy was absorbed in that tissue and converted into heat. Why this matters for the fetusImagine that a beam enters the uterus and passes through amniotic fluid. The fluid absorbs very little energy compared with many solid tissues. It can therefore transmit the wave toward the fetus with relatively little attenuation. When the wave reaches fetal tissue, some energy is reflected, some is scattered, and some is absorbed locally. If the beam intersects a fetal structure with higher absorption, local energy deposition can be greater than in the fluid surrounding it. This is one reason why we cannot simply calculate the total loss from the transducer to the fetus and then conclude that the remaining energy is harmless. We need to know where the energy is deposited, not just how much remains. F. The machine's output versus the acoustic field inside the motherNow we get to the distinction I think is central to your research. When an ultrasound manufacturer measures acoustic output, the measurements are generally made in water using a calibrated hydrophone. The FDA's guidance describes how those measurements are reported and how derated values are calculated to estimate output under a standardized tissue-attenuation assumption. U.S. Food and Drug Administration That means there are several different quantities we need to keep separate:
The machine's output is the starting point. It is not a direct measurement of fetal exposure. A normalized exampleLet's say the acoustic intensity at the transducer face is represented by 100 arbitrary units. Using our illustrative layered calculation: Illustrative intensity remainingNot a measurement from a real ultrasound machine or pregnancy At the transducer 100After skin, fat, muscle, uterus and fluid 0.40Approximately 0.4% of the starting intensity in this particular model. The actual acoustic field is more complicated than this simple calculation. For example:
That last point is particularly important. A B-mode image is constructed by sending pulses along many different lines. A particular fetal location is not necessarily exposed continuously while the image is being produced. Doppler modes can have different pulse characteristics and time-averaged outputs. The FDA's guidance distinguishes quantities such as spatial-peak pulse-average intensity and spatial-peak temporal-average intensity for precisely this reason. U.S. Food and Drug Administration+1 G. Why attenuation alone cannot tell us the temperature increaseWe have now estimated how much intensity remains at different depths. But we have not yet calculated temperature. To do that, we would need additional information:
A simplified way to express the relationship is: \[
\text{Absorbed power} \approx
\text{Incident power}\times\text{absorbed fraction}
\] But temperature rise is not simply proportional to absorbed power. Heat is being transferred and dissipated while it is deposited. This is why the Thermal Index (TI) is useful but should not be mistaken for a thermometer. It is an estimate related to potential heating under specified modeling assumptions, not a direct measurement of the temperature of the fetus. The FDA explicitly discusses acoustic output, thermal and mechanical indices, and their underlying calculations in its diagnostic ultrasound guidance. U.S. Food and Drug Administration H. What we can conclude so farThere are several important findings from following the beam physically. 1. Maternal tissue can substantially attenuate the beam. Skin, subcutaneous tissue, and especially muscle can contribute meaningful attenuation. 2. The amniotic fluid is a relatively efficient acoustic pathway. It does not attenuate the beam nearly as much as many solid tissues. 3. Reflection at ordinary soft-tissue boundaries is usually modest. A boundary involving bone can reflect substantially more energy. 4. The machine's measured output is not the same as fetal exposure. The actual field depends on the acoustic path, the beam, the frequency, and the operating mode. 5. We cannot calculate fetal heating from attenuation alone. We need local absorption, exposure duration, and heat-transfer information. And there is one particularly important research issue: the standard FDA derating model, the average effective attenuation measured in living humans, and a layer-by-layer model can produce very different estimates. That uncertainty deserves to be examined rather than hidden behind a single percentage. Sources worth keeping for this sectionThese are especially useful because they go beyond general patient information. 1. In-vivo ultrasonographic exposimetry: human tissue-specific attenuation coefficients in the gynecologic examination (1999). This is the most directly relevant human measurement study for our tissue-layer calculations. 2. Quantitative ultrasound imaging of soft biological tissues: a primer for radiologists and medical physicists (2021). A useful explanation of attenuation, frequency dependence, and the distinction between absorption and scattering. 3. FDA: Marketing Clearance of Diagnostic Ultrasound Systems and Transducers (2023). This is the primary regulatory reference for output measurements, derating, and acoustic-output indices. 4. Ultrasound — Medical Imaging Systems. A textbook-style explanation of propagation, attenuation, and the effect of frequency on penetration. 5. A Review on Biological Effects of Ultrasounds: Key Messages for Clinicians (2023). Includes tissue attenuation values and discussion of biological mechanisms. Where I would take the investigation nextI would make the next step more quantitative, but with a little more control over the assumptions. Rather than using one hypothetical pregnancy, let's build three models:
For each, we can calculate the acoustic intensity at the fetal surface and then at selected depths within fetal tissue. We can also compare B-mode and Doppler exposure assumptions. One important limitation: these will be modeled estimates, not measurements of actual fetal dose. To calculate a real device-specific estimate, we would eventually need the manufacturer's acoustic-output data for the transducer and operating mode. That should give us a much better foundation for the next question: what happens to the energy once it actually enters fetal tissue? Additional videos and textbook resources1. Video lectures — understanding the journey through tissueBritish Medical Ultrasound Society — Ultrasound Physics and Equipment Best overall resource King's College London · Dr. Colin Deane This is a structured university-level lecture series with videos and accompanying slides.
Ultrasound Reflection and Acoustic Impedance ETphysics · Approximately 20 minutes Particularly useful for understanding why some energy crosses a boundary while some is reflected. It works through reflection coefficients and transmitted intensity. Watch the videoAcoustic Impedance and Reflection Explained Radiology Tutorials · 2026 A shorter explanation of impedance differences and why they determine the amount of reflection at tissue boundaries. Watch the videoUniversity of Strasbourg — Ultrasound Tissue Interactions University educational module This one is especially relevant to our project. It includes a video, explanatory material, a PDF, and a quiz. It addresses transmission, reflection, absorption, scattering, and time-of-flight measurements. Open the educational module2. Textbooks and college-level readingThese are the resources I'd put alongside the written explanations.
Links:
3. The most important sections to connect to our calculationsI would attach these readings to the specific parts of the chapter:
One additional resource I particularly want to flagThe WHO manual is worth keeping close throughout this project. It is not simply a patient-information pamphlet. It is a technical teaching resource, and it explains the physical interactions between ultrasound and matter in a way that directly supports our calculations. I would use the BMUS lectures to see and hear the physics, the WHO manual to establish the physical principles, and the NCBI textbook chapter to work through the mathematics. That gives us three complementary ways of examining the same phenomenon rather than relying on one source or one explanation. Controversial claims and counterclaims: The dangers of prenatal ultrasoundThis is probably one of the most important chapters in the entire project, because there are legitimate scientific questions here that sometimes get mixed together with exaggerated claims. I want to approach it a little differently from a typical medical safety summary. For each controversy, we'll look at:
One distinction should remain central throughout: ultrasound is non-ionizing, but that does not mean it is biologically inert. It is mechanical energy. It can produce heating and mechanical effects in tissue. The question is whether the exposure produced during diagnostic pregnancy examinations causes meaningful harm to a developing human fetus. The FDA itself acknowledges the potential for heating and cavitation, while also describing diagnostic ultrasound as generally safe when used prudently by trained providers. FDA+1 1. Claim: Prenatal ultrasound can interfere with fetal brain developmentActive research question The concern One of the more serious arguments is that ultrasound waves could affect developing neurons, particularly during the period when neurons are migrating into their eventual positions in the cerebral cortex. This is a biologically interesting question because fetal brain development involves highly organized cell migration, differentiation, and formation of neural connections. The concern is not necessarily that ultrasound visibly damages brain tissue. It is that an exposure might produce subtle changes in development that are not apparent at birth. Evidence that prompted the concern Laboratory animal research has examined ultrasound exposure during fetal development, including studies involving neuronal migration and other developmental processes. There is also a human study that attracted considerable attention: - Stoltenberg et al./Grether et al. research group? More specifically, the 2018 JAMA Pediatrics paper by Abramowicz and colleagues examined prenatal ultrasound exposure and autism spectrum disorder. The study included 420 children and compared ultrasound exposure among children with autism, children with developmental delays, and typically developing children. The researchers did not find that the number of scans or total duration of exposure was greater in the autism group. However, they did find greater mean ultrasound penetration depth in the autism group compared with certain control groups. The authors suggested that this finding deserved further investigation. JAMA Network+1 The counterclaim A larger 2023 study using data from the CDC's Study to Explore Early Development examined 1,524 pregnancies. It found no association between the number of prenatal ultrasound examinations and autism diagnosis. The researchers also found no association in analyses examining individual trimesters. PubMed This is an important counterpoint because it investigated the same general concern using a larger sample and more detailed clinical information. My assessment The 2018 study raised a legitimate question, but it did not demonstrate that ultrasound causes autism. There are several limitations:
The 2023 study provides stronger evidence against a relationship between the number of scans and autism, although it cannot answer every question about acoustic output, frequency, intensity, and exposure duration. Bottom line: There is no established causal relationship between routine prenatal ultrasound and autism. The specific question of whether unusual acoustic exposures can influence fetal neurodevelopment remains appropriate for research. 2. Claim: Ultrasound can cause developmental delays, including speech and learning problemsThe concern Some studies have reported associations between prenatal ultrasound exposure and later childhood outcomes, including:
This is a broader concern than autism. It asks whether prenatal ultrasound could produce subtle developmental changes that might not become apparent until years later. Evidence behind the concern A 2002 review by Marinac-Dabic and colleagues summarized human epidemiological studies that had reported possible associations between prenatal ultrasound and outcomes such as growth restriction, delayed speech, dyslexia, and non-right-handedness. The authors emphasized that exposure measurements were inadequate in many studies and recommended better research incorporating acoustic output, exposure duration, timing, and confounding factors. PubMed A separate 2009 review examined 16 studies involving childhood development. Some studies reported small differences in particular outcomes, but the authors found methodological weaknesses and concluded that the evidence did not establish adverse developmental effects. PubMed The counterclaim The WHO systematic review examined 41 studies, including randomized trials, cohort studies, and case-control studies. It found no association between prenatal diagnostic ultrasound and:
It did report a weak association with non-right-handedness in boys, which I'll discuss separately. PubMed My assessment This is a good example of why we need to distinguish between a reported statistical association and an established biological effect. A small difference in a developmental test does not automatically mean ultrasound caused that difference. There are many potential confounding factors:
The available human evidence has not established that routine prenatal ultrasound causes speech delay, dyslexia, or learning disabilities. 3. Claim: Ultrasound increases the likelihood of left-handednessHistorical controversy This one is particularly interesting because it has been investigated for decades. The concern Some studies have reported that boys exposed to prenatal ultrasound were more likely to be left-handed or non-right-handed. The proposed explanation is that ultrasound might influence the development of the fetal nervous system, potentially affecting the development of brain lateralization. Brain lateralization refers to the specialization of certain functions in one hemisphere rather than the other. Evidence behind the concern The WHO systematic review found a weak association between prenatal ultrasound exposure and non-right-handedness in boys: \[
OR=1.26
\] With a 95% confidence interval of 1.03–1.54. That represents a modest statistical association, not proof of causation. PubMed Some Swedish registry studies also reported associations between prenatal ultrasound exposure and non-right-handedness in boys. The counterclaim The association has not been consistently demonstrated across studies. The Swedish Council on Health Technology Assessment reviewed the evidence and concluded that the available scientific evidence was insufficient to draw a reliable conclusion about a relationship between prenatal ultrasound and handedness. NCBI The WHO review also found no convincing evidence of broader neurological or intellectual impairment. My assessment This is not a demonstrated injury. It is a relatively small statistical signal that has not been consistently reproduced. Even if a relationship existed, left-handedness itself is not a disease or evidence of brain damage. The question is whether handedness is acting as a marker of some other neurological change. The available evidence does not establish that. 4. Claim: Repeated ultrasound examinations can restrict fetal growth or cause low birth weightImportant randomized trial This is one of the most important controversies because there is a randomized controlled trial behind it. The concern If ultrasound exposure can affect developing tissues, repeated exposure might influence fetal growth. This is not an unreasonable biological question. Fetal growth depends on many interacting systems, including placental function, blood flow, oxygen delivery, and cellular development. Evidence behind the concern: The Newnham trial, 1993 In this study, researchers randomly assigned 2,834 pregnant women to two groups:
The intensive group had a higher proportion of babies classified as small for gestational age. The reported relative risk for birth weight below the 10th percentile was: \[
RR=1.35
\] 95% confidence interval: \[
1.09-1.67
\] In other words, the study found a 35% relative increase in the risk of being classified as small for gestational age in the intensive-exposure group. PubMed This finding deserves attention. It should not be dismissed simply because later research has generally been reassuring. The counterclaim The same research group followed children from the trial up to eight years of age. The follow-up did not find significant adverse effects on childhood growth, neurological development, or intellectual outcomes attributable to the intensive ultrasound schedule. PubMed Another major randomized trial, the RADIUS study, enrolled 15,151 pregnant women and examined the effect of routine ultrasound screening on perinatal outcomes. It did not demonstrate an improvement in the major perinatal outcomes studied from routine screening, although that trial was primarily about screening effectiveness, not ultrasound bioeffects. New England Journal of Medicine My assessment The Newnham study is an important piece of evidence, but there are qualifications:
It does, however, support the principle that repeated exposure should not be assumed to be biologically irrelevant. This is one reason professional organizations recommend minimizing exposure time and using acoustic output prudently. 5. Claim: Ultrasound can heat fetal tissue enough to cause injuryEstablished physical mechanism; uncertain clinical risk at diagnostic exposures This is one of the strongest scientific concerns because ultrasound-induced heating is not hypothetical. It has been demonstrated experimentally. The concern As we discussed in the previous chapter, some acoustic energy is absorbed by tissue and converted into heat. Heating depends on:
Some fetal structures deserve particular attention. Bone absorbs ultrasound differently from soft tissue, and tissues near developing bone can experience localized heating. Evidence behind the concern Animal experiments have demonstrated temperature increases and biological effects under certain ultrasound exposure conditions. The AIUM's biological-effects statement reviews animal research and emphasizes that the effects depend on the ultrasound parameters and the tissue exposed. www.aium.org The FDA also recognizes the possibility of tissue heating during diagnostic ultrasound. FDA The counterclaim The existence of heating does not establish that ordinary diagnostic examinations produce harmful temperature increases. The ACOG guidance states that diagnostic ultrasound has not been associated with documented adverse fetal effects. It also explains that the potential for temperature elevation is generally lowest with B-mode imaging and higher with color and spectral Doppler. ACOG The distinction is between a possible physical effect and a harmful biological outcome. My assessment I would not frame this as "ultrasound does not heat the fetus." It can. The scientifically appropriate question is whether the temperature rise, in a particular tissue and for a particular duration, is sufficient to cause injury. The Thermal Index is useful for estimating potential heating, but it is not a direct measurement of fetal temperature. Nor does a TI of 1 mean that every tissue exposed has warmed by exactly 1°C. We should also remember that a brief temperature increase and a sustained temperature increase are not equivalent biological exposures. 6. Claim: Ultrasound can cause cavitation and mechanically damage fetal cellsThe concern Cavitation refers to the formation, oscillation, or collapse of gas bubbles in response to acoustic pressure changes. There are two commonly discussed forms:
In laboratory settings, cavitation can damage cells and tissues. This raises the question of whether diagnostic ultrasound could produce similar effects in a developing fetus. Evidence behind the concern Animal experiments have demonstrated mechanical bioeffects under certain conditions. Some experiments have found tissue damage, and research has demonstrated pulmonary capillary hemorrhage in animal lungs exposed to ultrasound. However, the lung is a special case because it contains gas-filled structures. The acoustic behavior of gas-containing tissue is very different from that of fetal soft tissue surrounded by fluid. AIUM reviews these findings and explains how biological effects depend on both the ultrasound source and the tissue being exposed. www.aium.org The counterclaim There is no established evidence that ordinary diagnostic ultrasound produces irreversible cavitation injury in human fetal tissue. A recent diagnostic-ultrasound safety review states that irreversible cavitation effects have not been independently confirmed in humans at modern diagnostic output levels. PMC+1 My assessment The existence of cavitation in laboratory experiments is not proof that it occurs in the fetus during a routine scan. We must be particularly careful not to transfer findings from experiments involving microbubble contrast agents or gas-containing tissues directly to a fetus in an amniotic-fluid environment. However, the absence of demonstrated injury is not the same as proof that every conceivable exposure is harmless. 7. Claim: Doppler ultrasound is more concerning than ordinary B-mode imagingImportant distinction between imaging modes This is a claim that has a substantial physical basis. The concern Doppler ultrasound is used to assess blood flow, including flow in the umbilical artery, uterine arteries, and fetal vessels. Unlike a conventional B-mode image, Doppler often repeatedly directs acoustic pulses into a smaller region to measure movement. Depending on the system and settings, this can produce a higher time-averaged acoustic output in the region being examined. Evidence behind the concern ACOG specifically identifies color and spectral Doppler as having greater potential for temperature elevation than B-mode imaging. ACOG The ISUOG guidelines for first-trimester Doppler examinations recommend particular caution with exposure duration and thermal index. The counterclaim Doppler is an important diagnostic tool. It can provide information about placental circulation, fetal anemia, fetal growth restriction, and other conditions. When medically indicated and performed using appropriate settings and exposure times, professional organizations consider its use acceptable. My assessment I would distinguish between two questions:
The timing matters, too. The earliest stages of embryonic development are not necessarily the same in terms of biological susceptibility as later pregnancy. This is why professional guidance treats early pregnancy Doppler with particular caution rather than assuming all ultrasound modes are interchangeable. 8. Claim: First-trimester ultrasound is more dangerous because the embryo is rapidly developingThe concern Early pregnancy is a period of rapid cellular division, tissue differentiation, and organ formation. The embryo's developing structures may be susceptible to environmental influences that would have different effects later in pregnancy. The concern is therefore that exposure during early development could have consequences that are not apparent immediately. Evidence behind the concern Animal studies demonstrate that developmental stage can influence the biological response to ultrasound. Some older evidence reviews have also pointed out that there is less direct human evidence for certain early-pregnancy exposures than for routine second-trimester examinations. The Swedish SBU review, for example, identified limitations in the evidence for first-trimester ultrasound and first-trimester Doppler. NCBI The counterclaim First-trimester ultrasound is clinically important. It can establish pregnancy location, estimate gestational age, assess viability, identify multiple pregnancies, and investigate possible complications. The fact that an embryo is developing rapidly does not demonstrate that diagnostic ultrasound causes developmental harm. My assessment There is a difference between saying that early development deserves extra caution and saying that early ultrasound has been shown to cause harm. The first statement is scientifically reasonable. The second is not established by the available evidence. This is a good example of why the ALARA principle is important: use the lowest acoustic output and shortest exposure that will answer the clinical question. 9. Claim: 3D and 4D ultrasound are more dangerous than conventional ultrasoundThe concern Three-dimensional ultrasound constructs a volumetric image, and four-dimensional ultrasound displays that volume over time. Some people assume that because these examinations involve more elaborate imaging, they must involve more exposure. That assumption is not necessarily correct. Evidence behind the concern The acoustic output depends on how the volume is acquired, the operating mode, the transducer, the settings, and how long the examination lasts. A 3D image may be acquired using conventional B-mode information, while some 4D imaging uses repeated real-time volume acquisition. The safety question is therefore not simply whether the image is 2D, 3D, or 4D. The counterclaim There is no established evidence that medically indicated 3D or 4D imaging causes fetal injury when performed appropriately. However, the evidence base for some specific exposure patterns is less extensive than for conventional diagnostic B-mode examinations. My assessment I would not automatically label 3D or 4D imaging dangerous. But I also would not assume that all 3D/4D examinations have identical exposure characteristics to a short conventional B-mode examination. Duration, output, and operating mode are more useful variables than the dimensional label. 10. Claim: Keepsake ultrasound examinations are an unnecessary riskThe concern Commercial keepsake ultrasound services may offer extended examinations for photographs, videos, or entertainment rather than a specific medical question. The concern is not necessarily that one additional image causes harm. It is that an examination may involve prolonged exposure without a corresponding medical benefit. Evidence behind the concern The FDA specifically discourages ultrasound performed solely for nonmedical purposes, including keepsake videos. It also warns about unsupervised fetal heartbeat monitors, which can lead to prolonged use or incorrect interpretation of findings. FDA The counterclaim A photograph or video captured during an otherwise medically indicated examination does not necessarily require additional exposure. The FDA itself distinguishes between obtaining keepsake images during a medical examination and conducting an additional examination solely for keepsake purposes. My assessment This distinction is important. A short video captured during a necessary scan is not the same thing as an extended commercial session designed to obtain numerous images. The question is whether the additional exposure is justified by a medical purpose. 11. Claim: Ultrasound machines became substantially more powerful after the 1992 regulatory changesA particularly important regulatory controversy This is one of the claims I think deserves its own investigation, because it is often presented without sufficient context. The concern In the United States, the permitted maximum derated spatial-peak temporal-average intensity for certain diagnostic ultrasound applications increased from approximately 94 mW/cm² to 720 mW/cm². That is an increase of approximately 7.7 times. \[
\frac{720}{94}\approx7.66
\] The concern is that modern machines may be permitted to operate at substantially higher acoustic output than older machines, while some of the reassuring human pregnancy research was conducted using older equipment. Evidence behind the concern A 2008 review by Douglas Miller, Safety Assurance in Obstetrical Ultrasound, discusses the history of acoustic-output limits and the changes in regulatory policy. Another review of fetal ultrasound mechanical effects points out that much of the human epidemiological evidence was collected using older diagnostic devices, with lower output limits. PubMed+1 This creates a legitimate research question: Can we confidently apply safety findings from older equipment to every modern exposure configuration? The counterclaim The increase in permitted maximum output does not mean that every ultrasound machine automatically operates at 7.7 times the previous intensity. It also does not mean that the fetus receives 7.7 times the acoustic energy. Actual exposure depends on the selected imaging mode, output settings, frequency, beam geometry, duration, and the tissue path. Modern systems also display the Thermal Index and Mechanical Index, allowing trained operators to monitor indicators related to potential bioeffects. My assessment I would not dismiss this concern. The increase in regulatory output limits is a real historical change, and it raises a legitimate question about how well older safety studies represent modern machines. But it would also be inaccurate to say that the FDA simply increased the permitted fetal dose by 7.7 times. The regulatory limit is not a measurement of actual fetal exposure, and it is not a guarantee that a machine will operate at that maximum. This is an excellent subject for a separate chapter examining the original FDA documents, the 1992 Output Display Standard, and the actual acoustic outputs of modern obstetric equipment. Putting the controversies side by sideHere is how I would summarize the current state of the evidence.
My overall interpretationI see three different levels of evidence here. 1. Established physical effects These are not controversial scientifically:
2. Human findings that deserve continued investigation These include:
These findings should not be ignored, but neither should they be treated as proof of injury. 3. Claims that have not been established There is no convincing human evidence demonstrating that routine, medically indicated prenatal ultrasound causes autism, intellectual disability, dyslexia, or widespread developmental damage. That is an important distinction. We can recognize genuine uncertainty without treating every proposed danger as established fact. Primary sources and research papers for this controversy chapterThese are the sources I would keep in the research bibliography. I've prioritized original studies and substantive reviews rather than relying entirely on professional organizations. 1. Newnham et al. (1993). Effects of frequent ultrasound during pregnancy: a randomised controlled trial. The Lancet. 2. Newnham et al. (2004). Effects of repeated prenatal ultrasound examinations on childhood outcome up to 8 years of age. The Lancet. 3. WHO (2009). Safety of ultrasonography in pregnancy: WHO systematic review of the literature and meta-analysis. Systematic review and meta-analysis 4. Abramowicz et al. (2018). Association of Prenatal Ultrasonography and Autism Spectrum Disorder. JAMA Pediatrics. 5. Pazol et al. (2023). Prenatal ultrasound use and risk of autism spectrum disorder: Findings from the Study to Explore Early Development. 6. AIUM (2021). Statement on Biological Effects of Ultrasound In Vivo. Animal research and biological mechanisms 7. Fetal ultrasound: mechanical effects. A review of mechanical bioeffects, animal experiments, and human evidence. 8. ACOG. Guidelines for Diagnostic Imaging During Pregnancy and Lactation. Professional clinical guidance 9. FDA. Ultrasound Imaging. FDA explanation of benefits, risks, and keepsake scanning 10. Swedish Council on Health Technology Assessment. Ultrasound for Fetal Assessment in Early Pregnancy. Evidence review and conclusions One final point I would not want us to loseThere is a tendency for discussions of medical safety to become polarized:
Neither argument is scientifically sufficient. The fact that a technology has a long history of use is reassuring, but it does not establish that every exposure pattern is harmless. Conversely, demonstrating a biological effect under experimental conditions does not establish that ordinary clinical use causes disease. The real question is exposure-specific risk. What frequency? What intensity? What mode? How long? At what stage of development? Which fetal tissue? What temperature change? What biological endpoint? General recommendations for prenatal ultrasoundFor an average-risk pregnancy, with no known complications or relevant medical history After looking at the physics, animal research, human studies, and professional guidelines, I would give the average pregnant woman a fairly straightforward recommendation: Have the medically recommended ultrasound examinations, but avoid unnecessary additional exposure. There is no good reason to be afraid of a properly performed diagnostic ultrasound, and no good reason to treat ultrasound as an entertainment activity that can be repeated without purpose. That position recognizes both sides of the evidence: diagnostic ultrasound has a long and generally reassuring safety record, but acoustic energy can produce biological effects under certain conditions. 1. What examinations are reasonable?For an uncomplicated pregnancy, a typical approach is:
The mid-trimester anatomy scan is recommended for all pregnant women by ISUOG. First-trimester scanning practices vary somewhat by country and healthcare system. Routine third-trimester scanning is not necessarily required for every uncomplicated pregnancy. Wiley Online Library+1 These are general patterns, not a prescription for every individual pregnancy. 2. How to minimize unnecessary exposureI would recommend the following:
This is the ALARA principle: As Low As Reasonably Achievable. It does not mean avoiding useful examinations. It means obtaining the necessary medical information without unnecessary exposure. ISUOG+2 3. Be more cautious about Doppler, particularly early in pregnancyNot all ultrasound modes have the same acoustic output.
This does not mean Doppler should be avoided when medically useful. However, during the early embryonic period, Doppler deserves particular care. ISUOG recommends limiting exposure and using Doppler only when clinically indicated. Its first-trimester guidance recommends keeping the displayed Thermal Index at or below 1.0 when Doppler is used between 11 and 14 weeks, with exposure generally limited to 5–10 minutes or less. ISUOG For an average pregnancy, there is no reason to request additional Doppler examinations solely for entertainment. 4. Avoid home fetal heartbeat monitorsI would put this fairly high on the recommendation list. Consumer fetal Dopplers can create two problems:
The FDA recommends that fetal heartbeat monitors be used by trained healthcare professionals when medically necessary. FDA 5. Don't be afraid of an additional medically necessary scanThis is an important counterbalance to everything we've discussed. If a healthcare provider recommends another ultrasound because of a concern about fetal growth, placental function, fetal position, or another medical issue, I would not recommend declining it simply because of theoretical ultrasound risks. The information gained may be important to the health of both mother and baby. ACOG considers ultrasound an appropriate imaging method during pregnancy when it is expected to answer a relevant clinical question or provide medical benefit. ACOG The goal is not to minimize the number of scans at all costs. It is to avoid scans that have no reasonable medical purpose. 6. What about keepsake and 3D/4D scans?My recommendation would be to skip separate commercial sessions performed solely to obtain photographs or videos. There is no established evidence that one brief keepsake image necessarily causes harm. The concern is unnecessary exposure, potentially prolonged scanning, and the absence of a medical reason for continuing. If 3D images are obtained during a medically indicated examination, that is a different situation. The dimensional format alone does not determine the exposure; the operating mode, settings, and duration matter. The FDA and AIUM discourage nonmedical fetal ultrasound examinations. FDA+1 7. What I would tell the average pregnant womanA practical ultrasound checklist8. My final assessment of the riskI would characterize the evidence this way:
The most important qualification is that absence of demonstrated harm is not proof of zero risk under every possible exposure condition. We do not have perfect long-term data for every modern machine, every imaging mode, and every possible exposure pattern. But we also have no persuasive evidence that the ordinary, medically indicated ultrasound schedule used in uncomplicated pregnancies creates a meaningful risk of developmental injury. 9. Sources for the general recommendationsThese are the sources I would use to support the concluding recommendations in the book:
Closing perspectiveIf I were reducing this entire investigation to one recommendation, it would be this: Use ultrasound as a medical diagnostic tool, not as a recreational activity. Have the examinations that provide meaningful information about the pregnancy, avoid unnecessary repetition, and don't refuse medically useful imaging out of fear of theoretical risks. That seems to be the position most consistent with both the physical science and the human evidence. And I think it avoids two mistakes: assuming that ultrasound must be completely harmless because it is not ionizing radiation, and assuming that it must be dangerous because biological effects can be demonstrated under experimental conditions. The evidence supports prudent use—not fear, and not unlimited use. |
