Best body scanner for a clinic: DEXA vs BIA vs 3D

Last updated: November 13, 2026

If you are buying a scanner for a med spa, plastics practice, wellness clinic, GLP-1 program, or nutrition practice, "best body scanner" is not a shopping list of bathroom scales.

This page is for the buyer, not the patient looking up a $20 Withings or Renpho.

The job is to pick the tool that can prove a treatment in the consult: fat loss, lean preserved, inches, a visual the patient will pay to see again. DEXA, professional BIA (InBody, Evolt), and optical 3D (ShapeScale, Styku, Fit3D) measure different physical things. They are not interchangeable.

What "best" means in a consult

A scale number cannot close a body-contouring or GLP-1 follow-up. Neither can a printout that moved because the patient was dry.

In a consult, that means:

• Error smaller than the 4-week change you are trying to show
• A result the patient believes next to a photo
• A retest the staff will actually run
• No radiation tax on every visit

DEXA is the reference method. It is a weak retest loop. Professional BIA is fast and cheap, and it loses the consult when hydration moves the number more than the treatment. Optical 3D is the class that sits in the room and shows shape plus composition.

The three methods

Seven devices show up on most clinic and gym shortlists. This is what each one is, using the makers' own published specs (checked 13 Nov 2026).

DEXA (dual-energy X-ray)

DEXA (DXA) sends two X-ray energies through the body and separates bone, fat, and lean from attenuation. It is the clinical reference most composition papers use, including ShapeScale V2 (GE Lunar iDXA).

It also measures bone density, which neither BIA nor optical 3D replaces. That is why endocrinology, research, and osteoporosis work still buy a table.

The cost of that reference is the workflow: ionizing radiation, a dedicated room, a trained operator, and a visit that does not look like a 60-second retest. You do not put a Lunar iDXA in the consult hallway and scan every GLP-1 follow-up.

Use DEXA when bone or a research-grade reference is the product. For aesthetics, it is a poor choice as the only proof tool.

Professional BIA (InBody, Evolt)

Professional BIA sends current through hand and foot electrodes and estimates fat, lean, water, and visceral fat from resistance. InBody (270 / 570 / 770) and Evolt 360 are the names on most clinic shortlists. Same measurement class: not DXA, not optical 3D.

The product is speed, not geometry.

The number moves with hydration, food, caffeine, alcohol, same-day training, and cycle phase. Both vendors say so. Evolt's knowledge base requires strict standardization. InBody's test-prep page tells operators not to scan until 3 hours after food and 1-3 days after exercise.

Evolt's accuracy page quotes "an average 98-99% repeatability, rivaling that of a DEXA scan." Repeatability is not agreement with DXA. A column that is consistently 4 points low is highly repeatable and still wrong for a consult. We did not put an Evolt 360 in V2, and we did not find a peer-reviewed Evolt 360 vs DXA paper with a named n / MAE.

Full write-up: Evolt 360 for clinics.

Independent papers on a professional InBody 770 vs DXA still show several-point error and a fat under-read:

• Potter et al., BMJ Nutrition 2022: InBody 770 vs GE Lunar iDXA in 795 fit Marines. Body-fat MAE 3.9 percentage points, bias -3.4.
• Lahav et al., European Journal of Clinical Nutrition 2021: InBody 770 vs DXA in 226 clinic patients (BMI 18-56). Mean BF% -3.1 pp, 95% limits of agreement +2.6 to -8.9.

That is the BIA ceiling, not a device scandal. A 4-week GLP-1 or contouring change often sits inside that noise. Full write-up: InBody scan accuracy for a practice.

Consumer smart scales (Withings, Hume, Renpho, Wyze, InBody Dial) are the same method with fewer electrodes. V2 ran an InBody Dial H20 against the same iDXA: body-fat MAE 5.01 pp, R2 0.44. That is a home trend tool, not a clinic scanner.

Optical 3D (ShapeScale, Styku, Fit3D)

Optical 3D captures surface geometry and estimates composition from the mesh, not from electrical resistance.

The class splits on capture:

• Rotating-platform systems (Styku, Fit3D ProScanner): the patient turns on a plate in front of a fixed camera. Circumference and a mannequin-style avatar are the usual output. Styku is the flagship comparison on this site; that page stays as-is: ShapeScale vs Styku.
• Tablet capture (Fit3D SNAP): the member turns 360 degrees in front of an iPad or Android tablet. Fit3D now calls its turntable ProScanner legacy and no longer recommends it for new installs.
• Rotating-head systems (ShapeScale): the scanner moves around a still patient and builds a photorealistic mesh. Composition is predicted from that geometry against DXA, not from a girth equation.

ShapeScale V2 (optical 3D vs GE Lunar iDXA, n=1,000+ adults, no participant overlap across train / validation / test):

• Body fat: MAE 1.89 pp, R2 0.91
• Lean mass (abstract): MAE 3.3 lb (1.49 kg), R2 0.98
• Appendicular lean mass: MAE 2.1 lb (0.97 kg), R2 0.96

V2 limitations are clothing, hair, and stance, not hydration. That is why the same scanner can sit next to a photo without the "you gained fat overnight" fight.

We did not publish a Fit3D-vs-iDXA MAE.

Do not treat any vendor "within 5%" line as one.

What that looks like in a consult

DEXA in a med spa or GLP-1 clinic shows up as a referral out, a radiation consent, and a PDF a week later. Useful as a baseline, weak as the thing you open on the iPad during the follow-up.

BIA shows up as:

• A patient who "gained fat" because they were dry
• A number that fights the photo
• Staff who stop trusting the device
• A printout the patient does not share

Optical 3D shows up as a body in the room, which is the product clinics actually sell.

DEXA vs BIA vs 3D (operator view)

The same four methods from the operator's side of the room.

How to choose

Start from what you have to show the patient or member, then pick the method.

Buy DEXA if the practice already bills bone density or needs a research reference. Pair it with something you can retest every 4 weeks. DEXA is the calibration. It is not the 4-week loop.

Keep professional BIA if you already run a tight protocol and only need a trend line. Do not buy another InBody or Evolt column if the bottleneck is trust in the consult. Detail: InBody scan accuracy and Evolt 360 for clinics.

Buy optical 3D if the next scanner has to sit next to a photo the patient believes. That is the plastics / med spa / GLP-1 / body-contouring job.

Do not buy a consumer smart scale for the practice. The old version of this page ranked Withings, Hume, Renpho, and Wyze. Those are home check-ins and cannot prove a treatment.

What each option costs (13 Nov 2026)

Published prices only. Where a maker quotes on a call, the table says so.

Book a ShapeScale demo

FAQ

What is the most accurate body scanner for a clinic?

DEXA is the reference method, and it is the only one of the three that measures bone density. Among scanners you can keep in the consult room, ShapeScale V2 came within 1.89 percentage points of a GE Lunar iDXA on body fat (mean absolute error) across more than 1,000 adults. Published InBody 770 studies against DXA show about 3 to 4 points of error (Potter et al., BMJ Nutrition, Prevention & Health 2022; Lahav et al., European Journal of Clinical Nutrition 2021).

How much room does a body scanner need?

ShapeScale needs a 5 x 5 ft space and a 7 ft ceiling. Styku lists a 6 x 4.3 ft minimum scanning area, Fit3D recommends 6 x 9 ft for the ProScanner, and GE shows the Lunar iDXA in an 11 x 10.6 ft exam room.

How long does a scan take?

A ShapeScale scan takes under a minute, and the patient stands still while the sensor moves around them. InBody puts a 770 test at 50 to 60 seconds. Fit3D says SNAP captures in under two minutes.

What does ShapeScale cost?

The 12-Month Plan is $299/mo billed monthly, or $199/mo billed annually ($2,388/yr), with a full-term warranty. To buy, it is $9,990 plus $199 delivery, with a 2-year warranty. Both come with a 45-day return period, starting on the delivery day.

Is a 3D body scanner a DEXA alternative?

Not for bone density, and it is not a DEXA scan. DEXA remains the reference method. ShapeScale estimates body fat, lean mass and appendicular lean mass from the 3D mesh, using a model trained on paired scans against a GE Lunar iDXA. Use DEXA when you need bone or a research baseline. Use optical 3D when the scanner has to live in the consult room and you want to retest often.

What is the difference between InBody, Evolt and a 3D scanner?

InBody and Evolt are bioelectrical impedance (BIA) devices. They send a small current through the body and estimate composition from resistance. They are fast, and many teams already know the report. An optical 3D scanner captures the body’s surface and shows a 3D model, circumferences and volumes. ShapeScale also estimates composition from that mesh. The methods are different, and they are not interchangeable.

What are the unique features of ShapeScale?

The person stands still while the sensor moves around them. You get a photorealistic 3D model, circumferences at the neck, shoulders, chest, biceps, waist, hips, thighs and calves, regional volumes, a volume heatmap, and composition estimates including appendicular lean mass, the lean tissue in the arms and legs. Staff review results on iPhone or iPad. The scan takes under a minute.

Does a body scanner belong in a gym, or only in a clinic?

Gyms, studios and clinics all buy this category. Keep a professional BIA device when your team already runs a tight composition protocol on a busy floor. Choose ShapeScale when trainers or clinicians also need a rotatable body, inches and regional volumes in the same review. DEXA usually needs a dedicated exam room and is a weaker weekly retest loop.

Sources

• ShapeScale V2 white paper (BF% MAE 1.89 pp, R2 0.91; lean MAE 1.49 kg / 3.3 lb, R2 0.98; ALM MAE 0.97 kg / 2.1 lb, R2 0.96; n=1,000+ vs GE Lunar iDXA; Dial H20 MAE 5.01 pp)
• Potter et al., BMJ Nutr Prev Health 2022, InBody 770 vs iDXA (MAE 3.9% BF, bias -3.4)
• Lahav et al., Eur J Clin Nutr 2021, InBody 770 vs DXA (-3.1 pp, LoA +2.6 to -8.9)
• Evolt, "How Accurate is the Evolt 360?" (98-99% repeatability)
• Evolt, scanning conditions
• InBody USA, test preparation
• Evolt 360 for clinics
• InBody scan accuracy for a practice
• ShapeScale vs Styku (Franco, Barr, Deutsch quotes)
• Resistanz Studio / Thurston Pym
• ShapeScale App Store (4.0 / 5, 43 ratings, iOS, 16 Aug 2026)
• Evolt Active App Store (1.9 / 5, 94, 16 Aug 2026)
• Evolt Active Google Play (2.7 / 5, 684)
• InBody app, US App Store (3.9 / 5, 1.8K, 16 Aug 2026)
• InBody app, Google Play (4.1 / 5, 12.9K, 16 Aug 2026)