The best of 19 handheld ultrasound probes tested on the Edinburgh pipe phantom scored a resolution integral of 66, a figure still lower than most cart-based cardiac and abdominal scanners built between 2010 and 2017 (Moran et al., Ultrasound in Medicine and Biology 2024). Center frequency sits behind most of the trade-offs in probe design: convex abdominal probes usually run at 2 to 8 MHz (Zander et al., Ultrasound International Open 2020), breast probes are expected to run at 12 MHz or higher (ACR Breast Ultrasound Accreditation 2026), and every added megahertz costs depth because soft tissue absorbs roughly 0.5 to 1.0 dB per centimeter per megahertz (Yao, Xia and Wang, Ultrasonic Imaging 2015). Regulators cap the energy that can be used to buy that depth back, at a derated ISPTA of 720 mW/cm² for most applications (FDA Marketing Clearance of Diagnostic Ultrasound Systems and Transducers 2023). We aggregated data from the American Society of Echocardiography, the American College of Radiology, and peer-reviewed phantom and physics studies in Ultraschall in der Medizin, The Ultrasound Journal, Diagnostics, Medical Physics and Insights into Imaging; for the commercial side of the same hardware, see our ultrasound transducer market statistics.
TL;DR
- Convex transducers for abdominal and pelvic work usually operate at 2 to 8 MHz (Zander et al., Ultrasound International Open 2020).
- Breast ultrasound transducers should run at a center frequency of at least 12 MHz, a requirement added on January 2, 2026 (ACR Breast Ultrasound Accreditation 2026).
- Pediatric echo labs need low-frequency (2 to 2.5 MHz) and high-frequency (above 7.5 MHz) transducers (ASE Pediatric TTE Guidelines 2024).
- Measured soft-tissue attenuation spans 0.01 dB/cm/MHz in blood to 4 dB/cm/MHz in muscle (Cloutier et al., Insights into Imaging 2021).
- Moving eye imaging from 10 MHz to 50 MHz improves axial resolution from 150 µm to 30 µm (Silverman, Clinical and Experimental Ophthalmology 2009).
- Phantom axial resolution across 7 scanners ranged from 0.6 to 1.9 mm, lateral from 1.3 to 8.7 mm (Herzog et al., Ultraschall in der Medizin 2024).
- FDA Track 3 limits: derated ISPTA up to 720 mW/cm² and MI up to 1.9, with ophthalmic use capped at 50 mW/cm² and MI 0.23 (FDA Diagnostic Ultrasound Guidance 2023).
- Lung capillary hemorrhage in laboratory mammals starts at an MI of approximately 0.4 (AIUM Statement on Biological Effects 2021).
- The top cardiac handheld scored a resolution integral of 44, versus 66 for the best small-parts handheld (Moran et al., Ultrasound in Medicine and Biology 2024).
- 66% of 35 POCUS experts would buy the Vscan Air today after rating six handhelds (Perez-Sanchez et al., The Ultrasound Journal 2024).
- About 100 handheld ultrasound devices are available worldwide (Dietrich et al., Diagnostics 2026).
1. Center Frequency by Clinical Application
Center frequency is the single biggest design decision in a probe, and the field has settled on bands rather than one number per exam. The spread is enormous: nominal transducer frequencies run from 1 to 25 MHz in general imaging (Zander et al., Ultrasound International Open 2020), and specialist eye and skin systems push far beyond that. The pattern is simple to read: the deeper the target, the lower the frequency.
Professional bodies now write frequency floors into their requirements. The ACR added an explicit 12 MHz minimum to its breast accreditation criteria on January 2, 2026, and notes that the center frequency is usually the value printed on the transducer (ACR Breast Ultrasound Equipment). The ASE asks pediatric labs to own several probes because a single frequency cannot serve both a newborn and a teenager (ASE Pediatric TTE Guidelines 2024).
| Metric | Value | Source |
|---|---|---|
| Typical operating range of convex probes used for abdominal and pelvic scanning | 2 to 8 MHz | Zander et al., Ultrasound International Open 2020 |
| Example center frequency of an abdominal or cardiac probe (with its bandwidth) | 2.5 MHz (1.25 to 3.75 MHz bandwidth) | Cloutier et al., Insights into Imaging 2021 |
| Low-frequency transducer a pediatric echocardiography lab should have | 2 to 2.5 MHz | ASE Pediatric TTE Guidelines 2024 |
| High-frequency transducer a pediatric echocardiography lab should have | Above 7.5 MHz | ASE Pediatric TTE Guidelines 2024 |
| Minimum center frequency for breast ultrasound transducers | At least 12 MHz, preferably higher | ACR Breast Ultrasound Accreditation 2026 |
| Conventional ophthalmic B-scan vs ultrasound biomicroscopy (UBM) of the eye | About 10 MHz vs 35 to 50 MHz | Silverman, Clinical and Experimental Ophthalmology 2009 |
| Skin imaging: frequencies that show dermis and epidermis vs epidermis only | 20 to 25 MHz vs 50 to 100 MHz | Barcaui et al., Anais Brasileiros de Dermatologia 2016 |
| Ultra-high-frequency skin imaging reported in a 2025 dermatology study | 70 MHz | Michelucci et al., Diagnostics 2025 |
Context: most recent available data for the ophthalmic and dermatology bands comes from reviews published in 2009 and 2016; the physics has not changed, and even by then a commercial 80 MHz scanner existed for imaging the drainage angle and Schlemm’s canal (Silverman, Clinical and Experimental Ophthalmology 2009). Catheter probes sit in between: capacitive micromachined (CMUT) intravascular arrays have been built at a center frequency of 20 MHz (Yu et al., Micromachines 2026), the class of device covered in our intravascular ultrasound statistics.
2. The Frequency vs Penetration Trade-off
Attenuation (the loss of signal strength as sound travels through tissue) is quoted per centimeter and per megahertz because it scales with both. A good rule of thumb for many soft tissues is 0.5 to 1.0 dB/cm/MHz, and the coefficient stays roughly proportional to frequency up to at least 300 MHz in skin (Yao, Xia and Wang, Ultrasonic Imaging 2015). The 0.5 figure is a convention, not a constant: measured values run from 0.01 dB/cm/MHz in blood to 4 dB/cm/MHz in muscle (Cloutier et al., Insights into Imaging 2021).
The arithmetic shows why abdominal probes stay low. At 0.5 dB/cm/MHz, a 3 MHz pulse reaching a target 15 cm deep travels 30 cm round trip and loses 0.5 x 3 x 30 = 45 dB. A 12 MHz pulse with the same 45 dB budget can travel only 45 / (0.5 x 12) = 7.5 cm round trip, which means a target depth of about 3.75 cm. That is our own illustration using the low end of the published rule of thumb, not a measured value.
| Metric | Value | Source |
|---|---|---|
| Rule-of-thumb attenuation coefficient for a variety of soft tissues | 0.5 to 1.0 dB/cm/MHz | Yao, Xia and Wang, Ultrasonic Imaging 2015 |
| Measured attenuation range, blood vs muscle (experimental and ex vivo extremes) | 0.01 vs 4 dB/cm/MHz | Cloutier et al., Insights into Imaging 2021 |
| Mean local attenuation in a liver example linked to stage 1 to 3 steatosis | 0.75 dB/cm/MHz | Cloutier et al., Insights into Imaging 2021 |
| Attenuation estimated in a simulated 0.5 dB/cm/MHz phantom at a 6 MHz transmit frequency | About 0.49 dB/cm/MHz | Omari and Varghese, Medical Physics 2014 |
| Imaging depth reached by 20 MHz skin probes | 8 to 10 mm (some manufacturers claim up to 15 mm) | Polanska et al., Frontiers in Medicine 2021 |
| Signal loss at the center frequency of an eye UBM after 12 mm to the focus and back | Over 13 dB | Silverman, Clinical and Experimental Ophthalmology 2009 |
| Depth of field, conventional 10 MHz eye probe vs 50 MHz UBM | 9.6 mm vs 0.85 mm | Silverman, Clinical and Experimental Ophthalmology 2009 |
Context: the Omari and Varghese phantom work (Medical Physics 2014) is most recent available data for validating the 0.5 dB/cm/MHz convention, and it was a simulation; real livers differ, which is why the Cloutier review treats measured attenuation as a marker of liver fat.
3. Measured Axial and Lateral Resolution in Phantom Studies
Axial resolution (the ability to separate two points lying along the beam) depends on pulse length, so it improves almost directly with frequency; it is typically about 1.5 mm at 1 MHz (Cloutier et al., Insights into Imaging 2021). Lateral resolution (separating points side by side) depends on beam width and focusing, and it degrades with depth. The 2024 Dresden phantom study shows the gap clearly: axial resolution stayed within 0.6 to 1.9 mm at every depth, while lateral resolution ranged from 1.3 to 8.7 mm and worsened with depth (r=0.6) (Herzog et al., Ultraschall in der Medizin 2024).
That study measured 100 µm pins placed 1 cm apart, at depths from 1 to 12 cm, on five handhelds and two cart workstations. At the high-frequency end, resolution moves from millimeters to tens of micrometers.
| Metric | Value | Source |
|---|---|---|
| Axial resolution range across 5 handhelds and 2 workstations (no correlation with pin depth) | 0.6 to 1.9 mm | Herzog et al., Ultraschall in der Medizin 2024 |
| Lateral resolution range across the same 7 devices (worsening with depth) | 1.3 to 8.7 mm | Herzog et al., Ultraschall in der Medizin 2024 |
| Reference cart workstation (GE Logiq E10): median lateral vs axial resolution | 5.33 mm (IQR 4.01 to 6.72) vs 1.21 mm (IQR 1.1 to 1.33) | Herzog et al., Ultraschall in der Medizin 2024 |
| Poorest median axial resolution among the 7 devices (Butterfly iQ+ handheld) | 1.6 mm | Herzog et al., Ultraschall in der Medizin 2024 |
| Eye imaging, axial and lateral resolution: conventional 10 MHz vs 50 MHz UBM | 150 and 450 µm vs 30 and 60 µm | Silverman, Clinical and Experimental Ophthalmology 2009 |
| Skin imaging, axial and lateral resolution: 20 MHz vs 50 MHz | 80 and 200 µm vs 20 and 100 µm | Polanska et al., Frontiers in Medicine 2021 |
| Axial resolution of a 20 MHz CMUT intravascular array, short pulse vs coded excitation | 60 µm vs 70 µm | Yu et al., Micromachines 2026 (citing Zangabad et al.) |
Context: phantom data on this scale is rare because it is slow to collect. The Edinburgh group’s pipe-phantom program, the largest open benchmark, covered 368 scanner and transducer combinations from 39 manufacturers over 15 years (Moran et al., Ultraschall in der Medizin 2022).
4. FDA Acoustic Output Limits for Diagnostic Transducers
Frequency choices sit inside a regulatory energy budget. ISPTA (spatial-peak temporal-average intensity) tracks heating potential, and MI (mechanical index) tracks the potential for mechanical effects such as cavitation; “derated” means the value is mathematically reduced to estimate the intensity inside tissue rather than in water. Ophthalmic scanning is held to an ISPTA of 50 mW/cm², about one fourteenth of the 720 mW/cm² allowed elsewhere (720 / 50 = 14.4), according to the FDA’s Marketing Clearance of Diagnostic Ultrasound Systems and Transducers guidance, issued February 21, 2023.
The preamendment levels (the application-specific reference levels of the FDA’s Track 1 pathway) show how the regulator ranks risk by application: fetal imaging was set at less than one seventh of the peripheral-vessel level (94 vs 720 mW/cm²).
| Metric | Value | Source |
|---|---|---|
| Track 3 global maximum derated ISPTA (most applications) | 720 mW/cm² or less | FDA Diagnostic Ultrasound Guidance 2023 |
| Track 3 global maximum mechanical output | MI of 1.9 or less, or derated ISPPA of 190 W/cm² or less | FDA Diagnostic Ultrasound Guidance 2023 |
| Thermal index (TI) above which the manufacturer must provide an explanation | 6.0 | FDA Diagnostic Ultrasound Guidance 2023 |
| Track 3 ophthalmic limits: TI, derated ISPTA and MI | TI 1 or less, 50 mW/cm², MI 0.23 | FDA Diagnostic Ultrasound Guidance 2023 |
| Preamendment derated ISPTA: peripheral vessel, cardiac, fetal imaging and other, ophthalmic | 720, 430, 94 and 17 mW/cm² | FDA Diagnostic Ultrasound Guidance 2023 |
Context: the guidance sets no general numeric ceiling on TI outside the eye; instead the on-screen indices put the decision in the operator’s hands, which is why the scan-time tables in section 5 matter.
5. Bioeffect Thresholds and Scan-Time Guidance
Regulatory caps are engineering limits; bioeffect thresholds are the biology underneath them. In laboratory mammals, lung capillary hemorrhage appears at an MI of approximately 0.4, far below the 1.9 regulatory ceiling, because gas bodies in the lung and intestine concentrate mechanical stress (AIUM, Statement on Biological Effects of Ultrasound in Vivo 2021). The same statement places the in vivo inertial cavitation threshold at approximately 3.6 MPa or more for frequencies of 0.5 MHz or greater.
For heat, the AIUM translates the on-screen thermal index into dwell-time limits, and recommends cutting those times by 33% for ARFI (acoustic radiation force imaging) and pulsed Doppler when bone is near the focus (AIUM, Recommended Maximum Scanning Times for Displayed TI Values 2022).
| Metric | Value | Source |
|---|---|---|
| Minimum MI threshold in laboratory mammals: lung capillary hemorrhage vs intestine | Approximately 0.4 vs 1.4 | AIUM Statement on Biological Effects 2021 |
| Maximum exposure at a 4.0 degree C temperature rise: fetal vs postnatal | 2 minutes vs 16 minutes | AIUM Statement on Biological Effects 2021 |
| Maximum exposure at a 1.5 degree C temperature rise: fetal vs postnatal | 1 hour vs 50 hours or more | AIUM Statement on Biological Effects 2021 |
| Obstetric and neonatal head or spine exams: TI with no time limit vs TI not recommended | 0.7 or less vs above 3.0 | AIUM Recommended Maximum Scanning Times 2022 |
| Abdominal, peripheral vascular and other non-eye exams: TI with no time limit vs TI not recommended | 1.5 or less vs above 6.0 | AIUM Recommended Maximum Scanning Times 2022 |
Context: these thresholds come from laboratory mammals and models, not from adverse events in patients, and the scan-time table was approved on October 30, 2016 and reapproved on September 26, 2022. Therapeutic systems that deliberately cross these thresholds are a separate category, covered in our HIFU statistics.
6. Handheld vs Cart-Based Systems: Phantom Benchmarks
The resolution integral (R) is a single figure of merit: the ratio of beam penetration to beam width, so a higher score means a probe sees both deep and sharp. On that scale, the best handheld for cardiac work scored 44, while the best small-parts handheld reached 66 (Moran et al., Ultrasound in Medicine and Biology 2024). Low-frequency phased arrays are where handhelds fall furthest behind: a 2026 consensus review notes that cardiac handheld R values fell below the lowest cart-based value, while small-parts values overlapped (Dietrich et al., Diagnostics 2026).
Phantom pins and clinical anatomy do not tell the same story. Handhelds matched or beat workstations on some phantom metrics, yet in ten volunteers the reference workstation still revealed more substructures (Herzog et al., Ultraschall in der Medizin 2024).
| Metric | Value | Source |
|---|---|---|
| Handheld devices benchmarked vs cart-based scanners (resolution integral) | 19 handhelds from 8 manufacturers (built 2016 to 2021) vs 142 cart-based scanners | Moran et al., Ultrasound in Medicine and Biology 2024 |
| Highest handheld resolution integral (Butterfly wired phased array, small-parts preset) | 66 | Moran et al., Ultrasound in Medicine and Biology 2024 |
| Highest handheld resolution integral for cardiac use (GE Vscan Extend phased array) | 44 | Moran et al., Ultrasound in Medicine and Biology 2024 |
| Cardiac handheld resolution integral values vs the lowest cart-based value | All below | Dietrich et al., Diagnostics 2026 |
| Liver portal-field substructures seen: reference workstation vs all handhelds combined (median) | 2 vs 1 (p<0.05) | Herzog et al., Ultraschall in der Medizin 2024 |
| Sigmoid colon wall layers seen: reference workstation vs Lumify vs Vscan Air (median) | 4 vs 2.0 vs 1.0 | Herzog et al., Ultraschall in der Medizin 2024 |
Outlier: in the same Dresden study, the sigmoid wall layers could not be identified at all with the Butterfly iQ+ in 5 of 10 subjects.
7. Handheld Probes: Expert Ratings and Device Landscape
Expert panels confirm what the phantoms show, and they also show that handhelds are not one category. In a 35-expert comparison of six devices, mean detail resolution scores ranged from 2.83 to 4.54 on a 5-point scale (Kruskal-Wallis p < 0.0001) (Perez-Sanchez et al., The Ultrasound Journal 2024). An average for handheld ultrasound as a category hides that spread, so device-level data matters more than category-level claims.
Transducer technology splits the field. Only the Butterfly iQ line uses CMUT chips in a clinically available handheld (Dietrich et al., Diagnostics 2026); its iQ uses a 9,000-element array, with a stated range of 1 to 10 MHz for the iQ and 1 to 12 MHz for the iQ3, figures that are the manufacturer’s own specifications as compiled by Yu et al. (Micromachines 2026).
| Metric | Value | Source |
|---|---|---|
| Detail resolution rating, Vscan Air vs Butterfly iQ+ (35 experts, 1 to 5 scale, mean) | 4.54 vs 2.83 | Perez-Sanchez et al., The Ultrasound Journal 2024 |
| Overall image-quality satisfaction, Vscan Air vs Butterfly iQ+ (mean) | 4.57 vs 2.37 | Perez-Sanchez et al., The Ultrasound Journal 2024 |
| Share of the 35 experts who would buy the Vscan Air today | 66% | Perez-Sanchez et al., The Ultrasound Journal 2024 |
| Detail resolution rating, Lumify vs Butterfly iQ+ (24 experts, 1 to 5 scale, mean) | 4.67 vs 3.17 (p<0.001) | Le et al., The Ultrasound Journal 2022 |
| Handheld ultrasound devices available worldwide | About 100 | Dietrich et al., Diagnostics 2026 |
| Scanning time per charge: conventional piezoelectric vs cMUT and pMUT handhelds | About 1 hour vs up to 2 hours | Dietrich et al., Diagnostics 2026 |
| PubMed-indexed handheld ultrasound publications: 1995 vs 2025 | 84 vs 2,124 | Dietrich et al., Diagnostics 2026 |
Context: in free-text comments from the same 35-expert study, 89% of remarks about the Butterfly iQ+ cited poor image quality, against 77% of Lumify remarks citing good image quality (Perez-Sanchez et al., The Ultrasound Journal 2024). The 2022 four-device study is most recent available data for that comparison set (Le et al., The Ultrasound Journal 2022).
Summary: Medical Ultrasound Transducer Center Frequency by the Numbers
| Metric | Value | Source |
|---|---|---|
| Nominal frequency range of general imaging transducers | 1 to 25 MHz | Zander et al., Ultrasound International Open 2020 |
| Convex abdominal and pelvic probes | 2 to 8 MHz | Zander et al., Ultrasound International Open 2020 |
| Minimum breast transducer center frequency | At least 12 MHz | ACR Breast Ultrasound Accreditation 2026 |
| Pediatric echo transducers: low vs high frequency | 2 to 2.5 MHz vs above 7.5 MHz | ASE Pediatric TTE Guidelines 2024 |
| Eye imaging: conventional B-scan vs UBM | About 10 MHz vs 35 to 50 MHz | Silverman, Clinical and Experimental Ophthalmology 2009 |
| Soft-tissue attenuation rule of thumb | 0.5 to 1.0 dB/cm/MHz | Yao, Xia and Wang, Ultrasonic Imaging 2015 |
| Measured attenuation, blood vs muscle | 0.01 vs 4 dB/cm/MHz | Cloutier et al., Insights into Imaging 2021 |
| Imaging depth of 20 MHz skin probes | 8 to 10 mm | Polanska et al., Frontiers in Medicine 2021 |
| Skin axial resolution, 20 MHz vs 50 MHz | 80 vs 20 µm | Polanska et al., Frontiers in Medicine 2021 |
| Phantom resolution across 7 scanners: axial vs lateral | 0.6 to 1.9 mm vs 1.3 to 8.7 mm | Herzog et al., Ultraschall in der Medizin 2024 |
| Track 3 derated ISPTA ceiling (non-ophthalmic) | 720 mW/cm² | FDA Diagnostic Ultrasound Guidance 2023 |
| Track 3 ophthalmic ceilings: ISPTA and MI | 50 mW/cm² and 0.23 | FDA Diagnostic Ultrasound Guidance 2023 |
| MI threshold for lung capillary hemorrhage in laboratory mammals | Approximately 0.4 | AIUM Statement on Biological Effects 2021 |
| Best handheld resolution integral, small parts vs cardiac (19 handhelds benchmarked against 142 cart-based scanners) | 66 vs 44 | Moran et al., Ultrasound in Medicine and Biology 2024 |
| Experts who would buy the top-rated handheld today | 66% of 35 | Perez-Sanchez et al., The Ultrasound Journal 2024 |
| Handheld devices available worldwide | About 100 | Dietrich et al., Diagnostics 2026 |
Methodology and Sources
- FDA, Marketing Clearance of Diagnostic Ultrasound Systems and Transducers (issued February 21, 2023; originally issued June 27, 2019): Track 3 global maximum, ophthalmic and preamendment acoustic output levels.
- AIUM, Statement on Biological Effects of Ultrasound in Vivo (approved August 19, 2021): MI and temperature-rise thresholds.
- AIUM, Recommended Maximum Scanning Times for Displayed Thermal Index Values (reapproved September 26, 2022): TI dwell-time tables.
- ACR, Equipment: Breast Ultrasound (revised January 2, 2026): 12 MHz center frequency floor.
- ASE, Guidelines for Performing a Comprehensive Pediatric Transthoracic Echocardiogram (Lopez et al., JASE, February 2024): pediatric transducer frequencies.
- Zander et al., Ultrasound Image Optimization (Knobology): B-Mode, Ultrasound International Open 2020: frequency ranges by probe type.
- Cloutier et al., Quantitative ultrasound imaging of soft biological tissues, Insights into Imaging 2021: attenuation ranges, axial resolution at 1 MHz, probe example.
- Yao, Xia and Wang, Multi-scale Functional and Molecular Photoacoustic Tomography, Ultrasonic Imaging 2015: 0.5 to 1.0 dB/cm/MHz rule of thumb.
- Omari and Varghese, Signal to noise ratio comparisons for ultrasound attenuation slope estimation algorithms, Medical Physics 2014: phantom attenuation estimate.
- Silverman, High-resolution ultrasound imaging of the eye: a review, Clinical and Experimental Ophthalmology 2009: ophthalmic frequencies, resolution, depth of field.
- Polanska et al., High-Frequency Ultrasonography in Teledermatology, Frontiers in Medicine 2021: skin imaging depth and resolution.
- Barcaui et al., High frequency ultrasound with color Doppler in dermatology, Anais Brasileiros de Dermatologia 2016: skin frequency bands.
- Michelucci et al., Prurigo Nodularis at Ultra-High-Frequency Ultrasound, Diagnostics 2025: 70 MHz skin imaging.
- Herzog et al., Technical assessment of resolution of handheld ultrasound devices and clinical implications, Ultraschall in der Medizin 2024: phantom and clinical resolution.
- Moran et al., Imaging performance of hand-held ultrasound scanners using the Edinburgh Pipe Phantom, Ultrasound in Medicine and Biology 2024: resolution integral benchmark (figures read from the indexed abstract).
- Moran et al., Edinburgh Pipe Phantom: 15 Years of Experience, Ultraschall in der Medizin 2022: scope of the phantom program.
- Dietrich et al., Handheld Ultrasound: Technical Review and International Expert Consensus Recommendations, Diagnostics 2026: device landscape, battery time, publication counts.
- Perez-Sanchez et al., Comparison of 6 handheld ultrasound devices by POCUS experts, The Ultrasound Journal 2024: expert ratings.
- Le et al., Comparison of four handheld point-of-care ultrasound devices by expert users, The Ultrasound Journal 2022: expert ratings.
- Yu et al., Advances and Challenges of CMUTs in Medical Imaging, Micromachines 2026: CMUT intravascular arrays and handheld specifications (vendor figures labeled as such).
- Data watch: the 0.5 dB/cm/MHz soft-tissue figure is a convention; Yao et al. give 0.5 to 1.0 dB/cm/MHz and Cloutier et al. report measured extremes of 0.01 to 4 dB/cm/MHz, so any depth estimate built on one number is approximate. The Dresden phantom paper gives the reference workstation’s median lateral resolution as 5.4 mm in its abstract and 5.33 mm in its results text; we used the results-text value. The skin resolution pairs (20 and 100 µm at 50 MHz, 80 and 200 µm at 20 MHz) are given with “respectively” after a passage defining axial before lateral resolution, so we read them as axial then lateral. Device weights conflict: the 2026 consensus review describes modern handhelds as 140 to 307 g, while the 2024 six-device comparison lists probes from 96 g to 309 g. Handheld market growth figures (a projected CAGR of up to 13.5% that the consensus review attributes to unnamed market analyses) were excluded because the underlying source is not disclosed. Several physics references (2009 to 2021) are most recent available data; the AIUM scan-time table and the FDA guidance are the documents most likely to be revised next.
Last updated: October 27, 2026. We update this roundup quarterly, and the next refresh is expected when the AIUM publishes its next review of the Recommended Maximum Scanning Times for Displayed Thermal Index Values statement.