Ultrasound-Guided Filler Injections: The Complete 2026 Guide

Quick Answer

Ultrasound-guided filler injection means imaging the treatment area with a high-frequency linear transducer before injecting — and in some workflows while injecting — so the operator sees the arteries and the existing filler instead of inferring them from surface landmarks. A 2025 modified Delphi consensus of 15 experts across seven specialties and 11 countries concluded that ultrasound imaging prior to injection is mandatory in areas at high risk for vascular adverse events. That is expert consensus guidance rather than randomized evidence showing improved patient outcomes.

The same consensus names the instrument. A minimum linear 15 MHz transducer is recommended for both learning and regular diagnostics, with B-mode, color Doppler, spectral Doppler and picture or video archiving all named as mandatory capabilities and power Doppler recommended (Velthuis et al., J Plast Reconstr Aesthet Surg 2025). The DERMUS Group reached the same 15 MHz figure nine years earlier (Wortsman et al., J Ultrasound Med 2016). Many of the mainstream handheld probes compared on this page publish maximum frequencies around 11–12 MHz or below by their own specifications, which places them well below that recommendation.

What guidance buys is measurable. In Teixeira’s 3-year retrospective series of 503 patients, pre-injection Doppler mapping took under 3 minutes and an artery larger than 0.8 mm was found sitting at the intended injection site in 34.7% of glabella cases (Teixeira, ASJ Open Forum 2025). When a vascular event does happen, ultrasound-guided targeted hyaluronidase resolved it with a mean of 122.5 IU against 1,519.4 IU for the blind flooding protocol (Schelke et al., Aesthetic Surgery Journal 2023).

What it does not buy is certainty. Across 318 published blindness cases with outcome data, 68.2% recovered no vision at all, and no treatment approach evaluated in those published cases reached statistical significance for visual improvement (Doyon et al., Aesthetic Surgery Journal 2024). DeLorenzi’s 2017 paper states the position plainly: the face has higher- and lower-risk areas, “but there are no zero risk areas.” Ultrasound reduces and manages risk. It does not eliminate it, and no scanner prevents blindness.

Start here · the whole field in one table

What the 15 MHz consensus recommendation leaves you, and what each scanner costs to own

You already know what a handheld wireless scanner is. The question an injector actually arrives with is narrower: where each of them sits against the frequency the published consensus recommends for this work, what it costs today, and whether the software keeps charging after the box is opened.

Two independent consensus documents recommend a minimum linear 15 MHz transducer with color Doppler. That one number orders the field, and it does it on published specification pages rather than on reputation. It orders rather than gates: frequency is a gradient, so the table states distance from the recommendation rather than a pass mark. Of the fourteen rows below, five publish maximum frequencies above the recommendation, one sits exactly at it, two sit one megahertz under it, five sit well under it, and one publishes no frequency at all. Everything below is each maker’s own published figure, with the kind of source it came from named in the cell.

Commercial disclosure, stated where it applies: Suresult publishes this guide and manufactures six of the scanners named in the table below. The table is organised around a figure Suresult did not write and cannot move — the minimum linear 15 MHz transducer recommended by two independent consensus documents — and every frequency in it was read from that manufacturer’s own published specification page. Prices carry the kind of source they came from, and a market range is never presented as an official figure. Two Suresult models publish 10/14 MHz and are shown sitting just below that recommendation, in the same row format and at the same level of detail as everyone else. The third column reports distance from a published purchasing recommendation, not a regulatory device classification. Nothing here is ranked, scored or recommended over anything else; the buyer-facing shortlist lives on a separate page, and the full conflict statement is in Methodology and disclosure.

Frequencies, depths and element counts are each manufacturer’s own published figures, read on July 28, 2026 — URLs and access dates in Sources. Every price carries the kind of source it came from: official store or official where the maker publishes a figure itself, market range where it does not and the configuration decides the number, dealer listings where only resellers quote it. A market range is a researched street figure, not an official price, and is not presented as one. Suresult figures are the currently displayed store prices; re-check before ordering. Clarius scanner prices and membership prices are listed separately on the official store, so the price column is the scanner alone and a five-year figure is the scanner plus whichever membership term is chosen. The L24 row carries the L24Pro photograph because this library holds no standalone L24 photograph, and the nearest alternative file is a sister-brand L24 that would name the wrong instrument.

The dermatologic systems that sit outside that table

One class sits deliberately outside that table. Dedicated dermatologic systems — the DUB SkinScanner at 22–75 MHz and the Longport EPISCAN at 20–50 MHz — are tabletop and console instruments rather than handhelds, so they are not a like-for-like row in that table. Both run far above the 15 MHz recommendation rather than near it, and neither publishes a price: they are quote-only, in a research and dermatology-clinic class of their own. These systems belong to a dermatologic ultrasound category and are not direct replacements for handheld point-of-care scanners; they sit here as context for why the 18–24 MHz handheld band exists at all. They resolve more and reach less, and a question about dermal architecture at tens of micrometres is not the question an injector is asking while mapping a facial artery before a syringe goes in.

The rest of this article is the evidence underneath that table: what the published record says goes wrong, what imaging changes about it, and where the 15 MHz number came from in the first place.

Shopping rather than learning the technique? Now that the published figures are on the table, the scanner shortlist for injectors is the aesthetic ultrasound buyer guide, the house shelf for this specialty is handheld ultrasound for aesthetics, skin-diagnostic imaging sits on handheld ultrasound for dermatology, and the membership arithmetic behind the incumbent scanner is in the Clarius subscription cost guide.

The quick list: four numbers that decide this

Published incidence estimateRoughly 1 vascular occlusion per 6,558 treatmentsThe only population-level figure in the literature: 0.015%, credible range 1:5,300 to 1:8,000, from 138,496 filler treatments in the Netherlands in 2016 against 44 vascular adverse-event patients over 25 months (Schelke et al., 2020).

The ceiling68.2% of published blindness cases recovered no vision at allOf 318 cases with outcome data, 6.0% recovered completely and 25.8% partially. No treatment approach evaluated in those published cases reached statistical significance for visual improvement; initial visual acuity was the significant predictor, P<.001 (Doyon et al., 2024).

The dose difference122.5 IU targeted against 1,519.4 IU floodingRoughly one-twelfth the enzyme, P=0.028. Of 39 patients given blind flooding first, 92.3% had no clinical improvement; treated afterwards under guidance, 100% improved immediately (Schelke et al., 2023).

The equipment recommendationA minimum linear 15 MHz transducerTwo independent consensus documents landed on the same number nine years apart: the 2025 aesthetic-injectables Delphi and the 2016 DERMUS guidelines. Color Doppler and spectral analysis are named in both. Both are expert consensus documents rather than randomized outcome trials.

The scale of the problem · historical review of published blindness cases

Why prevention is the only part of this that works well

Filler blindness is rare, catastrophic and largely irreversible once it happens. That combination is the whole argument for imaging before the needle goes in rather than reaching for a rescue protocol afterwards.

511Published cases, 1906–2023Including 365 new cases between September 2018 and March 2023 — the rate of publication is rising, not falling.

79.6%Caused by hyaluronic acidThe share of the new cases attributed to HA, the material most injectors use most days.

40.6%Highest-risk site: the noseFollowed by the forehead at 27.7% and the glabella at 19.0%.

1:6,558Occlusion incidence per treatmentAbout 0.015% of treatments, credible range 1:5,300 to 1:8,000 — rare enough to be invisible in one practice, common enough to be certain across a market.

Visual outcome in the 318 cases with outcome data

6.0%

25.8%

68.2%

6.0% recovered vision completely 25.8% partially 68.2% not at all

The finding that should govern every purchase decision on this page. In the same review, no treatment approach evaluated in the published cases — hyaluronidase, steroids or intra-arterial thrombolysis — reached statistical significance for visual improvement; the one significant predictor was initial visual acuity (P<.001). An earlier review of 48 cases published between January 2015 and September 2018 found hyaluronic acid responsible in 81.3%, with the nose (56.3%), glabella (27.1%), forehead (18.8%) and nasolabial fold (14.6%) leading the sites, and 20.8% recovering vision completely. Sources: Doyon VC et al., Aesthetic Surgery Journal 2024;44(10):1091–1104; Beleznay K et al., ASJ 2019;39(6):662–674; Schelke L et al., ASJ 2020;40(8):NP457–NP460. Accessed Jul 28, 2026.

The case for imaging

Why knowing the anatomy is not enough

Every injector learns facial vascular anatomy. The problem is that the textbook describes a population and the needle enters an individual, and Doppler ultrasonography has now quantified the gap.

A 2025 systematic review and meta-analysis of Doppler studies found the facial artery ran medial to the nasolabial fold — the type A, textbook course — in 46.2% of cases. It crossed medial to lateral (type C) in 22.5% and ran lateral (type B) in 12.0%. Mean depth was 6.27 mm at the mandibular border and 8.04 mm at the nasal ala, with mean diameters of 2.14 mm and 1.46 mm respectively, and Doppler visualization rates were essentially 100% at all three facial levels examined (Pourani et al., J Cosmet Dermatol 2025). Barely half of patients match the atlas, and Doppler is the modality that helps identify patient-specific vascular anatomy before injection.

The forehead makes the same point about depth rather than course. An ultrasound-based investigation found the deep branch of the supratrochlear artery changes plane from deep to superficial relative to the frontalis muscle at a mean of 14 mm from the superior orbital rim, and the supraorbital artery at a mean of 13 mm in males and 14 mm in females (Cotofana et al., Aesthetic Surgery Journal 2021). The “safe deep plane” of the forehead is only safe below a transition point that sits at a different height in every patient.

The glabella is worse. Measured at rest, the mean distance between the supratrochlear artery and the ipsilateral vertical glabellar frown line was 10.59 mm in males (SD 4.0, range 2.9–19.0 mm) and 8.21 mm in females (SD 4.0, range −3.3 to 14.2 mm) (Sigrist et al., Diagnostics 2024). That negative lower bound is not a rounding artefact: in some women the artery lies on the opposite side of the frown line from where the landmark predicts it.

What mapping finds when someone actually looks

The largest evidence of what pre-injection scanning changes is a 3-year retrospective series of 503 patients and 1,109 cc of hyaluronic acid injected under pre-injection Doppler mapping with a 20 MHz handheld scanner. An artery larger than 0.8 mm was sitting at the intended injection site in 34.7% of glabella cases (8 of 23), 19.2% of nose cases and 13.2% of nasolabial fold cases (43 of 324), with smaller shares at the chin, tear trough, marionette and temple. Measured diameters ran 0.8–1.8 mm for the facial artery (mean 1.2 mm at the nasolabial fold), 0.8–1.6 mm for the supratrochlear (mean 1.08 mm) and up to 1.8 mm for the superficial temporal. Within that series mapping took under 3 minutes, and there were zero vascular complications reported across it (Teixeira, Aesthetic Surgery Journal Open Forum 2025).

A narrative review of the same territory reframes the vocabulary usefully, describing vascular safe zones rather than danger zones (Cotofana et al., Plastic and Aesthetic Nursing 2022). Doppler is what turns a population-level safe zone into this patient’s safe zone.

How it is actually done

The two scanning workflows, and what each one asks of you

Published best-practice papers describe two accepted approaches, and they are not competing philosophies so much as different answers to how much time and how much hand-eye coordination a given clinic can spend (Vasconcelos-Berg et al., Diagnostics 2024).

Scan before injecting maps the region and marks the vessels on the skin immediately pre-procedure, then injects with the map in front of you. Scan while injecting holds the probe in one hand and the cannula or needle in the other, keeping the tip in view in real time. They answer different questions, and the difference decides which one a given clinic can afford.

Doppler is not optional in either. The vessels an injector needs to see are the internal carotid branches — supratrochlear, supraorbital and dorsal nasal — and the external carotid branches, superficial temporal and facial; pre-injection vascular mapping is recommended for the glabellar wrinkle, forehead, temple, nose and nasolabial fold, the relatively more dangerous filler sites (Lee, Archives of Plastic Surgery 2023).

That makes mapping a colour Doppler task at a particular depth rather than a general imaging task. The Pourani meta-analysis puts the facial artery at a mean of 6.27 mm at the mandibular border and 8.04 mm at the nasal ala, so a probe used for this has to resolve flow across that range with something in reserve beneath it. Suresult’s L20 and L20Pro publish 16/20 MHz with depth presets at 10/20/30/40 mm, which covers those depths and keeps a deeper window available; the buyer that fits is an injector who maps before every high-risk region and wants one probe for the face plus occasional body work. Suresult publishes this article and builds those probes — the full conflict statement is in Methodology and disclosure.

Filler safety is layered rather than solved. A structured Swiss cheese model proposes four prevention strategies stacked against each other, and imaging is one of them (Huang et al., J Cosmet Dermatol 2024):

1. Comprehensive facial anatomy knowledge. The scan is read by the anatomy you already know; a probe does not supply it.
2. Adherence to safe injection principles. Aspiration, slow low-pressure delivery, cannula selection and volume discipline sit here, untouched by imaging.
3. Use of real-time imaging and auxiliary instruments. This is where ultrasound sits — one layer of four.
4. Procedure checklists. The layer most clinics skip, and the cheapest to implement.

Ultrasound moves the odds. It does not remove the risk.

The rescue, and its limits

When something goes wrong: what guidance changes about the response

This section reports what published studies found. It is not a treatment protocol, and no dose here should be read as a recipe — management of a filler vascular adverse event belongs to a clinician’s own training, indemnity and local guidance.

The headline comparison is a 2023 study in the Aesthetic Surgery Journal. Ultrasound-guided targeted hyaluronidase resolved filler vascular adverse events with a mean of 122.5 IU (SD 34) against 1,519.4 IU (SD 1,137) for the blind regional flooding protocol, P=0.028 — roughly one-twelfth the enzyme dose. The sequence inside that study is the part worth sitting with: in 39 patients who first received flooding, 92.3% had no clinical improvement and 7.7% had only slight improvement; when those same patients were subsequently treated under ultrasound guidance, 100% showed immediate improvement (Schelke et al., 2023).

Three independent groups corroborate the direction. A 5-patient guided series reversed acute HA vascular occlusion completely in the 4 patients treated promptly, using a mean of 87 ± 44 IU (range 30–150 IU) with no early or late side effects (Urso et al., ASJ Open Forum 2024). A 2025 pictorial case series and systematic review using a 20 MHz linear probe reported 123 IU against the ~1,519.4 IU average required by the flooding method, with symptoms resolving within about 15 minutes in some cases and delayed administration associated with slower or incomplete recovery (Azizi, Wortsman et al., ASJ Open Forum 2025). And Doppler shows the endpoint: treating the exact site of obstruction and then re-imaging produced patent facial arteries and veins in 100% of patients in one reported cohort, with smaller doses sufficient because injection was precise (Munia et al., J Clin Aesthet Dermatol 2022).

What the images actually show

An international multicenter study of 100 patients with clinically diagnosed facial vascular adverse events after filler — four radiology, one dermatology and one plastic surgery centre, May 2022 to April 2025 — found the most common color Doppler findings were absent flow in perforator vessels (42%), absent flow in major vessels (35%), compensatory flow (26%), the string sign (18%) and increased peak systolic velocity (16%). Of the cohort, 93% were female, 98% had received HA fillers and 79% had already received hyaluronidase before the Doppler evaluation. The authors add a caution that belongs in any purchase decision: Doppler reference standards are well established for the carotids and the extremities, but no such reference standards exist for the face (Sigrist et al., RSNA 2025 abstract; peer-reviewed version in Diagnostics 2026).

Ultrasound also answers a question blind treatment cannot: whether the product is actually gone. An ultrasound-guided investigation of hyaluronidase acting on 11 different HA fillers found the largest volume reduction in the first hour — a 64.1% decrease from initial volume — reaching 81.7% total reduction at 24 hours, with efficacy varying by the filler’s elastic modulus, G′ (Bravo et al., J Cosmet Dermatol 2024).

The honest ceiling

None of this closes the case. The reference blind protocol, High Dose Pulsed Hyaluronidase, reported no partial or complete skin loss across two years of clinical use provided it was implemented within 2 days of the ischemic event — and in the same paper its author states that the face has higher- and lower-risk areas for filler treatment “but there are no zero risk areas” (DeLorenzi, Aesthetic Surgery Journal 2017). Murray et al. published UK consensus guidance on high-dose pulsed hyaluronidase for filler-induced vascular occlusion; published protocols vary, and management depends on a clinician’s own training and local guidance. The same document acknowledges that high-frequency ultrasound “gives the ability to visualize location, depth, and size of the filler embolus” and that ultrasound-guided treatment has required substantially lower hyaluronidase doses (Murray et al., J Clin Aesthet Dermatol 2021).

Set against the outcome data at the top of this page, the conclusion is uncomfortable but clean. Skin ischemia is frequently salvageable and the enzyme dose needed to salvage it drops by an order of magnitude under guidance. Vision loss is a different category: 68.2% of published cases with outcome data recovered nothing, and no treatment approach evaluated in those cases reached statistical significance for visual improvement. That asymmetry is why the consensus made pre-injection imaging mandatory in high-risk areas rather than making rescue imaging mandatory. No device on any market prevents blindness, and any vendor implying otherwise is selling past the evidence.

The diagnostic half of the job

Reading filler on ultrasound: the sonographic signatures

The second reason to own a scanner has nothing to do with injecting. Sonography has become the first-line imaging modality for detecting and identifying common cosmetic fillers, precisely because reliable information on a patient’s injection history is often difficult to obtain (Wortsman, J Ultrasound Med 2015). Every injector has met the patient who does not know what is in their face. The material has a signature whether or not the patient remembers it: the originating work established that fillers generate different patterns of echogenicity and posterior acoustic artefacts in skin and subcutaneous tissue (Wortsman et al., JEADV 2012).

Appearances as described in Tao Y, Wei C, Su Y, Hu B, Sun D. “Emerging High-Frequency Ultrasound Imaging in Medical Cosmetology.” Frontiers in Physiology 2022;13:885922 (open access). Clinical-consequence column is the authors’ framing, not the paper’s. Accessed Jul 28, 2026.

The identification is not theoretical. A high-frequency ultrasound study of 94 patients with filler in the nasolabial groove distinguished hyaluronic acid (35 cases), polyacrylamide (34), lipofilling (10), silicone oil (7), calcium hydroxylapatite (6) and silicone implants (2), concluding that echo intensity, echo characteristics, boundary clarity, fluidity and blood-flow signals can be used as the basis for identifying filler materials (Jiang et al., Aesthetic Plastic Surgery 2022). For clinics that want to report findings consistently, a standardized nomenclature for sonographic description and reporting of soft tissue fillers was proposed after review of 39 published articles, identifying ten parameters for describing and monitoring deposits (Schelke, Cassuto, Velthuis and Wortsman, J Cosmet Dermatol 2020).

The bridge to equipment

The equipment specification: what the consensus actually recommends

This is the part of the topic where marketing has run ahead of the literature, so it is worth stating the recommendation before naming a single scanner. Two independent consensus processes, nine years apart, arrived at the same recommended instrument specification.

The 2025 Delphi was a 15-expert, four-round process across seven medical specialties and 11 countries; it also recommends ultrasound-guided injection in many parts of the face for fillers, toxin and injection lipolysis. A third document, the WFUMB Position Paper on best practice in aesthetic dermatologic ultrasound (Chammas, Sigrist, Wortsman et al., Ultrasound Med Biol 2025;51(11):2173–2193), issued 24 consensus recommendations covering equipment, technique, documentation and guided procedures; its individual recommendations sit behind a paywall and are cited here for scope, not quoted. All three are expert consensus documents rather than randomized outcome trials. Accessed Jul 28, 2026.

Why 15 MHz, and why 20–24 MHz for the face

The 15 MHz figure is a physics statement, not a preference. Penetration and what a frequency can resolve trade against each other in a fixed way: 10 MHz penetrates about 35 mm and shows epidermis, dermis and subcutaneous tissue; 20 MHz penetrates about 10 mm and resolves epidermis, dermis and part of the subcutis; 50 MHz penetrates 3–4 mm and shows epidermis and dermis; 100 MHz reaches about 1.5 mm and sees epidermis only (Mlosek et al., Journal of Ultrasonography 2020). A general-purpose 10–12 MHz probe images deep enough for almost any body-cavity question and cannot resolve the dermal and subdermal detail a filler-plane decision depends on. Depth of visualization runs the same way: roughly 50 mm at 8 MHz against roughly 10 mm at 20 MHz (Lee, 2023).

Published protocols land above that figure rather than on it. Best-practice papers describe broadband linear probes spanning 6–24 MHz, 4–20 MHz and 6–18 MHz, with 18–20 MHz used for detailed dermal and subdermal anatomy and 24 MHz reserved for the superficial subcutaneous plane (Vasconcelos-Berg et al., 2024). The 503-patient mapping series ran on a 20 MHz handheld, and the 2025 hyaluronidase series used a 20 MHz linear probe. Fifteen megahertz is the recommended minimum; the working range in the literature is 18–24 MHz.

Frequency behaves as a gradient rather than a cliff. A probe quoted at 6–24 MHz spends part of its range below the recommended minimum and part well above it, and because wavelength scales inversely with frequency, 14 and 15 MHz resolve superficial tissue almost identically — a few per cent apart. The distance from the recommendation is what matters, and at 11 or 12 MHz that distance is several times larger and bites hardest at exactly the shallow depths filler work occupies.

Where each scanner sits, by its own published specifications

The full reading sits near the top of this page: the table of published frequencies, prices and software terms covers every scanner an injector is likely to be offered, each figure carrying the source it was read from. Two observations belong with that table, and they land better here, after the physics.

The first is that the 15 MHz figure orders the field rather than closing it. None of the mainstream point-of-care handhelds in that table reaches it, because many mainstream handheld systems publish frequency ranges aimed at broader point-of-care applications; what separates them from a 14 MHz linear probe is not a shared verdict but distance, and the distance is several times larger. The second is that a depth preset is not a resolution claim. A 16/20 MHz probe offering a 40 mm display window is showing you 40 mm; Mlosek’s table says the frequency resolves roughly the first 10 mm of it with the detail that matters for a filler plane. Both are true at once, and a spec sheet quoting only the first is answering half the question.

Where the house line sits, stated plainly

Commercial disclosure, restated where it applies: the six scanners named in the paragraphs that follow are built by Suresult, which also publishes this article. Two of the six publish 10/14 MHz and are named as sitting just below the recommended minimum, and that recommendation comes from consensus documents Suresult had no hand in writing. Full statement in Methodology and disclosure.

Four listed Suresult models sit above the published recommendation — the L24, L24Pro, L20 and L20Pro — and two sit just below it, the L14 and L14Pro. The honest version of that sentence is more useful to an injector than a pitch.

Take the requirement first: published protocols reserve 24 MHz for the superficial subcutaneous plane and use 18–20 MHz for detailed dermal and subdermal anatomy. The L24 and the L24Pro publish 18/24 MHz with depth presets of 10/15/20/25 mm — 192 elements on the L24, 256 elements and a four-hour battery on the L24Pro — so on published specification the band and the depth window are the ones those papers describe. The buyer that fits is an injector working the dermal and immediately subdermal planes, tear-trough and superficial filler review included, rather than deep structural work.

The L20 at 16/20 MHz covers the 18–20 MHz band those papers use for detailed dermal and subdermal anatomy while keeping deeper presets for work that is not face-only, and the L20Pro carries that same band on a 256-element array. Each of those four publishes B-mode, color Doppler, M-mode, power Doppler and pulsed wave Doppler, and export to JPEG, MP4, PNG and DICOM — the Delphi’s mandatory capability list, archiving included, item by item. Those capabilities are read from each model’s own published specification page rather than from a family-level claim.

For work below the superficial planes the frequency drops by design, and that is where the L14 and L14Pro sit — just below the figure this article is built on. Both publish 10/14 MHz: the L14 over depths of 20–80 mm on a 192-element array, the L14Pro on depth presets of 20/40/60/80 mm and a 256-element array. That is one megahertz under the recommended minimum — close enough that 14 and 15 MHz resolve superficial tissue almost identically, and far enough that neither is the model the consensus documents describe when they name a minimum. Stated plainly: for dermal filler review these are the wrong two to buy, and the L14Pro product page positions the family the same way, naming musculoskeletal medicine among its uses and describing it as the deeper linear option. An injector buying primarily for facial mapping is better served by one of the four models above the recommendation; MSK-adjacent and deeper soft-tissue work is the case mix these two are built for. The table states the distance rather than issuing a verdict, which is the only reason the table is worth reading.

Three things that change the five-year number

Frequency decides how close a scanner sits to the published recommendation. These three decide what owning it is like, and an injector tends to meet them after the purchase order.

1. Whether the software renews. The incumbent 20 MHz handheld is sold at $5,395 for the scanner, with membership priced separately at $695 a year or $2,500 one-time, as of July 2026; the five-year figure is the scanner plus whichever membership term is chosen. Suresult scanners carry lifetime free software with unlimited scans on iOS, Android and Windows and no subscription cost, so choosing a scanner at or above the recommended frequency is not also choosing a renewal to re-approve every year. Across five years that gap is real money on hardware that images identically on day 1,825.
2. Whether the paperwork exists before you ask for it. A prescription imaging device in a med-spa attracts questions from landlords, insurers and inspectors that a laser never did. Suresult keeps FDA, ISO, MDR/CE, TGA and ANVISA records in a public documentation library rather than releasing them per request, which turns a two-week email chain into a link.
3. Whether you can send it back. Scanning a real face on a real clinic day is the only test that predicts whether a workflow survives a schedule. Suresult ships every scanner with a 30-Day Worry Free Trial, so that test happens on your own patients rather than at a booth.

The operator, not the instrument

Training and competence: what the same consensus says about you

Both consensus documents that set the recommended equipment specification also address the operator, and the second recommendation is harder to buy than the first. The DERMUS guidelines put a number on maintaining competence: a suggested minimum of 300 examinations per year, alongside routine color Doppler and spectral curve analysis (Wortsman et al., 2016). The 2025 Delphi recommends the 15 MHz minimum explicitly for learning as well as for regular diagnostics — the recommendation is stated for both in the same sentence.

Independent training exists and is not endorsed here in either direction. Structured courses are offered by academies and by several of the authors cited on this page; some are device-neutral and some are run with a manufacturer, and an injector should establish which before paying. The sequence most clinics arrive at is: identify the vessels reliably in a low-stakes setting, map before injecting on the high-risk regions only, extend to the rest of the face, then consider real-time guided injection. Nothing in the literature supports skipping to the last step.

One caution from the multicenter Doppler study belongs here too: reference standards for facial Doppler do not exist the way they do for the carotids and extremities. An operator learning this is learning to recognise patterns, not to check a number against a published cut-off, and that is a slower kind of competence to acquire.

Quick answers

Ultrasound-guided filler questions, answered

The same measure for everyone

Methodology and disclosure

Commercial disclosure: Suresult manufactures six of the scanners named in the equipment table — the L24, L24Pro, L20, L20Pro, L14 and L14Pro — and this article is published on Suresult’s site. We handle that conflict the only honest way we know: the reference figure in this article is a published consensus recommendation we did not write and cannot influence, applied identically to every scanner in the table and reported as distance rather than as a pass mark; two of the six house models sit just below that recommendation and are shown sitting below it, in the same row format as everyone else; no scanner is ranked, scored, awarded a position or recommended over another anywhere on this page, and readers looking for a shortlist are sent to a separate guide; every frequency and price in the table was read from that manufacturer’s own published page, with each access date recorded in Sources; and every clinical claim is attributed to a named peer-reviewed paper by authors with no relationship to this publication.

We built this article standard-first. The clinical spine is 30 claims drawn from peer-reviewed sources published between 2012 and 2026, each with a date and a verifiable URL, and no figure appears in the text that does not trace to one of them. Where a paper reports a mean and a range, both are given; where it reports a P value, the P value is given; where a review is narrative rather than quantitative, it is cited for framing and not for numbers. The WFUMB position paper is cited for its existence and scope only, because its 24 individual recommendations sit behind a publisher paywall and we do not quote what we have not read in full.

Device facts are limited to published specification claims: frequency, depth presets, element counts, imaging modes, export formats, software terms and price. No device here is credited with improving a clinical outcome, because no source on this record supports that link for any brand, house included. Two things were deliberately left out: any vendor marketing claim about reduced hyaluronidase volumes, since the studies attach their findings to a technique rather than to a product; and any attribution of the 2025 multicenter Doppler percentages to a named instrument, since that study spanned six centres and did not report one. Prices are the figures each maker displayed on the access date and should be re-checked before ordering; regulatory wording follows the FDA’s own distinction, in which 510(k) devices are cleared rather than approved.

Nothing on this page is clinical advice, a treatment protocol or a substitute for training. Hyaluronidase figures are reported as published study findings with their citations, and management of a filler vascular adverse event belongs to a clinician’s own training, indemnity cover and local guidance.

Sources & verification dates

• Doyon VC, Liu C, Fitzgerald R, Humphrey S, Jones D, Carruthers JDA, Beleznay K. “Update on Blindness From Filler: Review of Prognostic Factors, Management Approaches, and a Century of Published Cases.” Aesthetic Surgery Journal 2024;44(10):1091–1104. DOI 10.1093/asj/sjae091 — academic.oup.com/asj/article/44/10/1091/7649223. Accessed Jul 28, 2026.
• Beleznay K, Carruthers JDA, Humphrey S, Carruthers A, Jones D. “Update on Avoiding and Treating Blindness From Fillers: A Recent Review of the World Literature.” Aesthetic Surgery Journal 2019;39(6):662–674. PMID 30805636 — pubmed.ncbi.nlm.nih.gov/30805636/. Accessed Jul 28, 2026.
• Schelke L, Decates T, Kadouch J, Velthuis P. “Incidence of Vascular Obstruction After Filler Injections.” Aesthetic Surgery Journal 2020;40(8):NP457–NP460. PMID 32538425 — academic.oup.com/asj/article/40/8/NP457/5857552. Accessed Jul 28, 2026.
• Schelke LW, Velthuis PJ, Decates T, Kadouch J, Alfertshofer M, Frank K, Cotofana S. “Ultrasound-Guided Targeted vs Regional Flooding: A Comparative Study for Improving the Clinical Outcome in Soft Tissue Filler Vascular Adverse Event Management.” Aesthetic Surgery Journal 2023;43(1):86–96. PMID 35951759 — pubmed.ncbi.nlm.nih.gov/35951759/. Accessed Jul 28, 2026.
• Urso SU, Molinari P, Fundarò S, Mosti G. “Use of Minimal Amounts of Hyaluronidase in the Ultrasound-Guided Treatment of Acute Vascular Occlusion by Hyaluronic Acid: A Preliminary Report.” ASJ Open Forum 2024;6:ojae025. PMID 38938923 — pubmed.ncbi.nlm.nih.gov/38938923/. Accessed Jul 28, 2026.
• Azizi N, Tootoonchi N, Khorasanizadeh F, et al., Wortsman X. “Ultrasound-Guided Hyaluronidase Injections for the Management of Filler-Induced Arterial Ischemia: A Pictorial Case Series and Systematic Review of Literature.” ASJ Open Forum 2025;ojaf125. PMCID PMC12586333 — pmc.ncbi.nlm.nih.gov/articles/PMC12586333/. Accessed Jul 28, 2026.
• Munia MA, Munia CG, Parada MB, Parente JB-H, Wolosker N. “Doppler Ultrasound in the Management of Vascular Complications Associated with Hyaluronic Acid Dermal Fillers.” J Clin Aesthet Dermatol 2022;15(2):40–43. PMCID PMC8884183 — pmc.ncbi.nlm.nih.gov/articles/PMC8884183/. Accessed Jul 28, 2026.
• Teixeira DA. “A New Time-Saving Technique for Ultrasound-Guided Filler Injection: A 3-Year Retrospective Study.” ASJ Open Forum 2025. DOI 10.1093/asjof/ojaf061 — academic.oup.com/asjopenforum/article/doi/10.1093/asjof/ojaf061/8163485. Accessed Jul 28, 2026.
• Pourani MR, Ebrahimzade M, Goudarzi E, Kassir M, Robati R, Moravvej H, Abdollahimajd F. “Evaluation of Facial Artery Course Variations, Diameters, and Depth Using Doppler Ultrasonography: A Systematic Review and Meta-Analysis.” J Cosmet Dermatol 2025;24(9). PMID 40874402 — pubmed.ncbi.nlm.nih.gov/40874402/. Accessed Jul 28, 2026.
• Cotofana S et al. “The Change of Plane of the Supratrochlear and Supraorbital Arteries in the Forehead—An Ultrasound-Based Investigation.” Aesthetic Surgery Journal 2021;41(11):NP1589–NP1598 — academic.oup.com/asj/article-abstract/41/11/NP1589/6154313. Accessed Jul 28, 2026.
• Sigrist R, Desyatnikova S, Chammas MC, Vasconcelos-Berg R. “Best Practices for the Use of High-Frequency Ultrasound to Guide Aesthetic Filler Injections—Part 1: Upper Third of the Face.” Diagnostics 2024;14(16):1718. PMCID PMC11353849 — pmc.ncbi.nlm.nih.gov/articles/PMC11353849/. Accessed Jul 28, 2026.
• Vasconcelos-Berg R, Desyatnikova S, Bonavia P, Chammas MC, Navarini A, Sigrist R. “Best Practices for the Use of High-Frequency Ultrasound to Guide Aesthetic Filler Injections—Part 2: Middle Third of the Face, Nose, and Tear Troughs.” Diagnostics 2024;14(22):2544. PMCID PMC11593187 — pmc.ncbi.nlm.nih.gov/articles/PMC11593187/. Accessed Jul 28, 2026.
• Lee W. “Hyaluronic Acid Filler Injection Guided by Doppler Ultrasound.” Archives of Plastic Surgery 2023;50(4):348–353. PMCID PMC10411166 — pmc.ncbi.nlm.nih.gov/articles/PMC10411166/. Accessed Jul 28, 2026.
• Velthuis PJ, Alfageme F, Cartier H, et al., Wortsman X. “The use of ultrasound imaging in aesthetic injectables: A modified Delphi consensus.” J Plast Reconstr Aesthet Surg 2025;104:33–37. PMID 40101353 — pubmed.ncbi.nlm.nih.gov/40101353/. Accessed Jul 28, 2026.
• Wortsman X, Alfageme F, Roustan G, et al. “Guidelines for Performing Dermatologic Ultrasound Examinations by the DERMUS Group.” J Ultrasound Med 2016;35(3):577–580. PMID 26887446 — pubmed.ncbi.nlm.nih.gov/26887446/. Accessed Jul 28, 2026.
• Chammas MC, Sigrist R, Alfageme F, et al., Wortsman X. “WFUMB Position Paper: Consensus on Best Practice in Aesthetic Dermatologic Ultrasound.” Ultrasound Med Biol 2025;51(11):2173–2193. PMID 40866164 — pubmed.ncbi.nlm.nih.gov/40866164/. Cited for scope; individual recommendations paywalled. Accessed Jul 28, 2026.
• Mlosek RK, Migda B, Migda M. “High-frequency ultrasound in the 21st century.” Journal of Ultrasonography 2020;20(83):e233–e241, Table 1. PMCID PMC7830059 — pmc.ncbi.nlm.nih.gov/articles/PMC7830059/. Accessed Jul 28, 2026.
• Tao Y, Wei C, Su Y, Hu B, Sun D. “Emerging High-Frequency Ultrasound Imaging in Medical Cosmetology.” Frontiers in Physiology 2022;13:885922, open access — frontiersin.org/journals/physiology/articles/10.3389/fphys.2022.885922/full. Accessed Jul 28, 2026.
• Wortsman X. “Identification and Complications of Cosmetic Fillers: Sonography First.” J Ultrasound Med 2015;34(7):1163–1172. PMID 26112618 — pubmed.ncbi.nlm.nih.gov/26112618/. Accessed Jul 28, 2026.
• Wortsman X, Wortsman J, Orlandi C, Cardenas G, Sazunic I, Jemec GBE. “Ultrasound detection and identification of cosmetic fillers in the skin.” JEADV 2012;26(3):292–301. PMID 21418333 — pubmed.ncbi.nlm.nih.gov/21418333/. Accessed Jul 28, 2026.
• Jiang L, Yuan L, Li Z, Su X, Hu J, Chai H. “High-Frequency Ultrasound of Facial Filler Materials in the Nasolabial Groove.” Aesthetic Plastic Surgery 2022;46(6):2972–2978. PMID 35699759 — pubmed.ncbi.nlm.nih.gov/35699759/. Accessed Jul 28, 2026.
• Schelke LW, Cassuto D, Velthuis P, Wortsman X. “Nomenclature proposal for the sonographic description and reporting of soft tissue fillers.” J Cosmet Dermatol 2020;19(2). PMID 31456355 — pubmed.ncbi.nlm.nih.gov/31456355/. Accessed Jul 28, 2026.
• Schelke LW, Decates TS, Velthuis PJ. “Ultrasound to improve the safety of hyaluronic acid filler treatments.” J Cosmet Dermatol 2018;17(6):1019–1024. PMID 30084182 — pubmed.ncbi.nlm.nih.gov/30084182/. Accessed Jul 28, 2026.
• Schelke LW, Velthuis P, Kadouch J, Swift A. “Early ultrasound for diagnosis and treatment of vascular adverse events with hyaluronic acid fillers.” J Am Acad Dermatol 2023;88(1):79–85 (online-first 2019). PMID 31325548 — pubmed.ncbi.nlm.nih.gov/31325548/. Accessed Jul 28, 2026.
• DeLorenzi C. “New High Dose Pulsed Hyaluronidase Protocol for Hyaluronic Acid Filler Vascular Adverse Events.” Aesthetic Surgery Journal 2017;37(7):814–825, source of “there are no zero risk areas.” PMID 28333326 — pubmed.ncbi.nlm.nih.gov/28333326/. Accessed Jul 28, 2026.
• Murray G, Convery C, Walker L, Davies E. “Guideline for the Management of Hyaluronic Acid Filler-induced Vascular Occlusion.” J Clin Aesthet Dermatol 2021;14(5). PMCID PMC8211329 — pmc.ncbi.nlm.nih.gov/articles/PMC8211329/. Accessed Jul 28, 2026.
• Bravo BSF, Cavalcante T, Silveira C, Bravo LG, Zafra MC, Elias MC. “Resolve and dissolve—An ultrasound-guided investigation on the effects of hyaluronidase on different soft tissue fillers.” J Cosmet Dermatol 2024;23(10):3173–3181. PMID 38769647 — pubmed.ncbi.nlm.nih.gov/38769647/. Accessed Jul 28, 2026.
• Sigrist R et al. “Doppler Ultrasound Findings in Filler-Related Facial Vascular Adverse Events: An International Multicenter Study” — RSNA 2025 press abstract, rsna.org/-/media/files/press/2025/2622/sigrist-abstract.pdf; peer-reviewed version Diagnostics 2026;16(11):1587. Accessed Jul 28, 2026.
• Cotofana S, Lowrey N, Frank K, et al. “Vascular Safe Zones for Facial Soft Tissue Filler Injections.” Plastic and Aesthetic Nursing 2022;42(4):238–245. PMID 36469395 — pubmed.ncbi.nlm.nih.gov/36469395/. Narrative review, cited for framing. Accessed Jul 28, 2026.
• Huang Y-L, Chi C-C, Chang S-L, et al. “A structured approach with Swiss cheese model to reduce vascular adverse events of filler injections.” J Cosmet Dermatol 2024;23(3):737–745. PMID 37864302 — pubmed.ncbi.nlm.nih.gov/37864302/. Accessed Jul 28, 2026.
• Clarius. L20 HD3 and L15 HD3 scanner pages and official US store (8–20 MHz, 4 cm depth, 25 mm field of view, $5,395; 5–15 MHz, 7 cm depth, $3,795; membership $695/year or $2,500 one-time) — clarius.com and store.clarius.com. Accessed Jul 30, 2026.
• Butterfly Network. iQ3 specifications page (“Frequency Range: 1-12 MHz”, preset depths 4–30 cm) and official store pricing: $3,899, software $299–$420 per year or $1,500 one-time, which Butterfly scopes to five years of software access — butterflynetwork.com/iq3-specs and the Butterfly store. Accessed Jul 28, 2026.
• GE HealthCare. Vscan Air support specifications (linear array 3–12 MHz, centre frequency 7.7 MHz, max depth 8 cm) and official Vscan Air pricing from $4,999, no subscription published — vscanair-support.gehealthcare.com and gehealthcare.com. Accessed Jul 28, 2026.
• Philips. Lumify L12-4 linear array transducer product page (“12 to 4 MHz extended operating frequency range”, 34 mm aperture); Philips publishes no list price for the L12-4, so the $5,995–$9,995 figure is a configuration-dependent market range compiled from reseller listings and is labelled as such in the table, with Reacts tele-ultrasound optional at $84 per user per year — usa.philips.com. Accessed Jul 28, 2026.
• EchoNous. Kosmos comprehensive specification sheet (Lexsa 128-element linear, approximately 4–11 MHz, centre 7.5 MHz); EchoNous publishes no list price, so the $7,975–$11,979 figure is a configuration-dependent market range compiled from reseller listings and is labelled as such in the table — echonous.com. Accessed Jul 28, 2026.
• Mindray TE Air specifications as described in a peer-reviewed device report (1.8–4.5 MHz phased array; no linear variant) — pmc.ncbi.nlm.nih.gov/articles/PMC10894582/. Pricing is from dealer listings rather than a manufacturer page and is labelled as such in the table: i3P $5,422, e5M $4,125, no forced subscription, with some tools sold as one-time upgrades. Accessed Jul 28, 2026.
• Vave Health. Universal probe page and FAQ — no operating frequency published on either — official store price $2,799 with no subscription — vavehealth.com/universal. Accessed Jul 28, 2026.
• taberna pro medicum. DUB SkinScanner applicator range 22–75 MHz, ~42 µm axial resolution at 38 MHz, 16 mm maximum imaging depth; quote-only — tpm-online.de. Accessed Jul 28, 2026.
• Longport Inc. EPISCAN I-200/I-700 broadband 20–50 MHz, ~40 µm vertical resolution, imaging to ~20 mm; quote-only — longportinc.com/episcan-i-200. Accessed Jul 28, 2026.
• Suresult product pages for the L24, L24Pro, L20, L20Pro, L14 and L14Pro (frequency, depth presets, element and channel counts, imaging modes, exported file formats, software terms, warranty, and prices as displayed: L20 $3,000 against a $4,800 list, L24 $3,700 against a $6,900 list, L20Pro $4,000, L24Pro $4,500, L14 $2,250 against a $3,200 list, L14Pro $3,700; the L14Pro page names musculoskeletal medicine among its uses) — suresultmed.com/shop/handheld-ultrasounds/. Specifications accessed Jul 28, 2026; prices re-read Aug 3, 2026.
• Suresult software, returns, warranty and certification pages (“No Subscription Cost. Lifetime Free Software, Unlimited Scans”; free software for iOS, Android and Windows; JPEG, MP4, PNG and DICOM export with local storage; the 30-Day Worry Free Trial; 1-year standard warranty with +2 year $195 and +3 year $375 extensions; FDA, ISO, MDR/CE, TGA and ANVISA records in the public documentation library) — suresultmed.com. Accessed Jul 28, 2026.
• Hero image: color Doppler capture of a superficial cutaneous vessel at 20 MHz, 20 mm depth, acquired June 5, 2026 on the house software interface and shown uncropped; header identity fields are demonstration entries. In-body sonograms: a 20 MHz cross-sectional color Doppler capture of a subcutaneous artery and its paired vein, and a 20 MHz B-mode skin-layer capture, both acquired June 5, 2026 on the house software interface and both skin and subcutaneous rather than facial captures; header identity fields are demonstration entries. Videos: Suresult official YouTube channel — a facial scanning demonstration published Jul 31, 2025, and “L20 Facial Ultrasound for Aesthetics” covering temple and superficial temporal artery imaging, both embedded Jul 2026 and accessed Jul 29, 2026. Scanner thumbnails: product photographs held in this publication’s own media library, one per row, each showing the model named in that row, except the L24 row, which carries the L24Pro photograph because the library holds no standalone L24 photograph.