SNAP by Fit3D | Clinical & Wellness Resources
Understanding body composition. Measuring meaningful change.
A practitioner's guide to the science behind SNAP by Fit3D, what body-scan results mean, how to collect reliable measurements and how to monitor meaningful change.
Using Fit3D scans in practice
A body scan is most useful when it changes a conversation from “What does the scale say?” to “What is actually changing, and what should we do next?”
SNAP by Fit3D uses a guided tablet scan to create a visual 3D model and provide body measurements and model-estimated composition metrics. It supports baseline assessment, monitoring and education in fitness, wellness and clinician-led weight-management settings.
This kit explains how to read results, standardize the scan, identify unreliable measurements and interpret change responsibly. It is intended for healthcare and wellness professionals, including those supporting GLP-1 treatment programs, nutrition coaching and fitness transformation.
Understand the data
Know what comes from the scan surface, what is inferred, and what is calculated.
Get a reliable baseline
Repeatable clothing, pose, framing and input checks improve confidence in trends.
Read progress responsibly
Interpret fat, lean, waist, weight and functional measures together.
Know when to verify
Use a repeat scan or another assessment when the finding is implausible or high stakes.
Practitioner quick reference
The most useful points to have beside you during a baseline or follow-up consultation.
Waist and other circumferences
Use the same anatomical landmark and method each time. These measurements describe the outer body surface, not internal fat compartments.
Body fat % and visceral fat
Useful for assessment and monitoring with appropriate caution. These are modeled estimates, not direct DXA, CT, or MRI readings.
Fat mass and lean mass
Computed from body weight and estimated body-fat percentage. A stale weight or inaccurate estimate affects both values.
Trends over single readings
Consider weight, waist, fat and lean trends together. Verify surprising changes before altering a treatment or coaching plan.
Before trusting a scan
- Correct person and current height, weight and profile details.
- Close-fitting clothing, no shoes, hair and facial hair clear of neckline.
- Full body visible, good lighting, complete 360° rotation and head facing straight.
- Comparable clothing type and scan conditions to previous assessments.
How SNAP by Fit3D works
From a guided tablet capture to body models, circumferences and composition estimates.
- Prepare the setting and client. Use an approved Samsung tablet with SNAP, a clear floor area, even lighting, suitable close-fitting clothing and accurate profile details.
- Capture a guided rotation. The client stands in the guided position and rotates through a full 360° while the tablet captures a short video, typically about 10 seconds.
- Reconstruct body shape. Computer vision processes multiple views to estimate a consistent digital 3D representation of the body's visible exterior.
- Extract measurements and estimate composition. The system identifies modeled body landmarks for circumferences and applies statistical models to estimate body-composition indicators.
- Review, verify and compare. The Fit3D results experience displays body models and metrics; staff review scan quality before using the results for a consultation or comparison.
Inputs that influence the report
Images and body shape
Clothing, pose, hair, lighting and rotation affect the body geometry the system can reconstruct.
Height and weight
Height helps scale measurements. Current body weight is used when translating estimated composition percentages into mass values.
Age and sex-related inputs
These may affect reference comparisons and model outputs; verify profile information and use the product's available options accurately.
Algorithm version
Models may evolve. Track versions and product notices when comparing results over long periods or interpreting published validation.
SNAP is the Fit3D tablet product. It should not be confused with legacy ProScanner hardware or with every device configuration supported by the underlying technology provider.
Back to top ↑Where the numbers come from
Understand the difference between surface-derived measurements, statistical estimates and calculations.
| Result | How to interpret it | What changes or limits it |
|---|---|---|
| 3D body model Reconstructed | A digital representation of the observed body surface, useful for visual side-by-side comparisons. | Clothing, occlusions, hair and reconstruction quality; visualization is not direct internal imaging. |
| Circumferences Surface-derived | Waist, hips, chest, arms, legs and related body dimensions sampled from the 3D model at defined landmarks. | Landmark definitions may differ from a manual tape protocol. Compare like with like. |
| Waist-to-height ratio Calculated | Waist circumference divided by height in matching units. A complement to broader risk assessment, not a diagnosis. | Highly dependent on the waist landmark and accurate height. |
| Body-fat percentage Model-estimated | Estimated proportion of body mass attributed to fat, inferred from body shape and model inputs. | Not directly measured fat tissue; errors may be larger in certain body types. |
| Fat mass Derived | Body weight multiplied by estimated body-fat fraction. | Inherits uncertainty from body-fat % and the entered weight. |
| Lean mass Derived | Body mass other than fat; includes water, organs, bone and muscle, not muscle alone. | Hydration shifts and composition-model uncertainty can affect interpretation. |
| Appendicular lean / skeletal muscle Model-estimated | Modeled indicators related to limb lean tissue or skeletal muscle; where displayed, assess alongside strength and function. | Not directly imaged muscle. Do not diagnose sarcopenia using SNAP alone. |
| Visceral fat Model-estimated | Estimate related to adiposity around abdominal organs, derived from external features and learned relationships. | Not CT or MRI volume measurement; avoid equating outputs from different methods. |
| BMR and metabolic age Derived indicators | Modeled calorie needs or comparative age-like communication metrics where available in the SNAP report. | Neither substitutes for indirect calorimetry, metabolic testing or individualized treatment. |
Two calculations every practitioner should know
For example, a person at 90 kg and 30% estimated body fat has an estimated 27 kg fat mass and 63 kg lean mass. If the weight field is wrong, the kilogram results will be wrong even if the percentage estimate stays the same.
Accuracy, precision and validation
Understanding what published comparisons show and what they do not prove.
Three questions to ask about any study
- Accuracy: How closely does a scanner estimate agree with a reference such as DXA or a calibrated 3D scanner?
- Repeatability: How closely do repeated measurements agree when the same person scans under comparable conditions?
- Applicability: Are the device, algorithm version, population and capture protocol comparable to the SNAP experience used in your practice?
Published findings from the underlying Prism technology
A Prism Labs technical report published April 28, 2026 describes 550 participants and 3,721 scans, with a DXA subset of 273 participants. In its reported adaptive body-fat model analysis, correlation with DXA was r = 0.950 and mean absolute error was 3.24 percentage points. These values are Prism-reported research figures, not a separate Fit3D-specific validation study. [1]
3.24 percentage points
Prism-reported mean absolute error for body-fat % against DXA, adaptive model analysis. It is not a maximum error or an individual guarantee.
r = 0.950
Reported correlation with DXA in the analysis. Correlation demonstrates association, not interchangeability or absence of bias.
0.27–0.81 cm
Range of mean within-person standard deviations for selected repeat circumference measures reported across body sites.
One patient ≠ study mean
Individual deviations may exceed published average errors, particularly when body type or protocol differs from the validation cohort.
Earlier third-party investigations and peer-reviewed work have also evaluated smartphone-derived body measures and body composition; see the evidence list for original publications and methodology. [2] [3]
How to explain it to a client
“The underlying scanning method has been compared with established reference methods in published research. Like every body-composition method, the reading has uncertainty. We will use a consistent process and combine the results with your weight, measurements, symptoms and program goals rather than treating a single estimate as exact.”
Back to top ↑Where scanning works well and where caution is needed
Know which differences are biological, which are technical, and when another assessment is appropriate.
| Situation | Possible effect | Recommended response |
|---|---|---|
| Loose clothing or a different clothing type | Altered apparent body surface, potentially affecting measurements and estimated composition. | Rescan in minimal or consistently form-fitting clothing; consider resetting baseline if conditions changed. |
| Hair or beard obscuring the neckline | Neck and upper-body contour can become less reliable. | Secure hair fully above neckline and position beard away from body contour where practicable. |
| Incorrect current weight or height | Errors in calculated mass estimates and dimensional scaling. | Verify weight with a consistent scale immediately before scanning and confirm height/profile. |
| Incomplete turn, tilted head, altered arm pose | Missing views or inconsistent geometry. | Follow app guidance and repeat a clean full 360° rotation. |
| Variable lighting, cluttered background or obstruction | Imperfect visual reconstruction. | Use clear space and bright, even lighting. |
| Very lean, highly muscular, or atypical body composition | External shape may be a less reliable predictor of internal composition. | Prioritize direct circumference trends and consider DXA or other assessments for high-stakes decisions. |
| Pregnancy, amputation, substantial edema or major shape alterations | Evidence may not generalize; anatomy and fluid status can alter assumptions. | Do not presume validated body-composition accuracy; select a suitable clinical assessment method. |
| Single unexpected change | Measurement variation, fluid shifts or a protocol mismatch may be mistaken for tissue change. | Review inputs, repeat under matched conditions, and seek confirmation over time. |
Do not use SNAP as a standalone diagnosis
SNAP outputs do not diagnose obesity-related disease, sarcopenia, malnutrition or metabolic disorders. In medical care, combine them with history, examination, lab work, validated clinical criteria and other appropriate methods. Eligibility, dosage, diagnosis and treatment decisions belong to licensed clinicians.
Standard clinical scanning protocol
A repeatable operator-led workflow for baseline and follow-up visits.
Before the client arrives
- Confirm approved Samsung tablet, installed SNAP app, power and connection.
- Prepare a private, uncluttered area with even lighting and enough room to rotate safely.
- Follow Fit3D's in-app placement and framing instructions for the specific tablet.
- Explain the image capture and scan-data privacy process and obtain appropriate consent.
Client preparation
For the most consistent body outline, use minimal or close-fitting clothing: men commonly wear briefs or tight boxer briefs without a shirt; women commonly wear close-fitting underwear or leggings with a sports bra. Adapt for privacy and comfort without forcing exposure. Avoid loose shirts, jackets, bulky shorts and shoes. Tie hair fully above the neckline; clear facial hair from the neck contour when feasible.
Ask clients to use the same clothing category at every visit. For better trend comparisons, aim for the same time of day and similar pre-scan conditions. Where practical, obtain weight just before scanning on the same scale and verify the result is entered correctly.
Operator workflow
- Confirm identity and profile. Check correct account, height, weight and relevant demographic inputs.
- Check clothing and hair. Ensure garment edges and hair will not distort body contour.
- Frame the entire body. Follow the SNAP on-screen placement instructions; keep the client visible throughout the movement.
- Guide one smooth full turn. Maintain the recommended arm position, keep head straight and rotate 360° at the pace indicated by the app.
- Inspect scan quality. Review the resulting 3D views for obvious artifacts or missing segments before interpreting numerical results.
- Document the session. Record date, relevant clothing changes, weight source, and unusual conditions that may influence comparison.
Scan validity checklist
- Form-fitting scan clothing was worn.
- Hair and beard, if applicable, were clear of the neckline.
- The user completed a full 360° turn while keeping their head straight.
- The model has no obvious geometry artifacts and profile/weight details are correct.
Cadence
Choose a cadence appropriate to the program. Weekly or biweekly visits can support engagement, while monthly scanning may produce easier-to-interpret body-composition changes for many programs. Avoid reacting to small changes between closely spaced scans when they may fall within normal measurement variation.
Back to top ↑Reading change over time
Use several connected signals rather than a single favorable number.
A useful review starts with the original goal and compares body weight, waist and body-fat estimates, fat mass, lean mass, body shape, and available functional or clinical indicators. Favor persistent patterns over single-visit swings.
Weight decreases, waist decreases and fat estimate declines
Baseline: 100 kg, 35% estimated body fat, 105 cm waist. Follow-up: 94 kg, 32% estimated body fat, 99 cm waist. Estimated fat mass changes from 35.0 kg to 30.1 kg; estimated lean mass from 65.0 kg to 63.9 kg.
Interpretation: The combined pattern is compatible with mostly fat-related weight reduction. Because composition is modeled, validate trends with standardized scans and other program information. Reinforce consistent nutrition, activity and follow-up.
Weight decreases but the lean estimate falls substantially
Baseline: 100 kg, 35% estimated body fat. Follow-up: 94 kg, 34% estimated body fat. Estimated fat mass changes from 35.0 to 32.0 kg; estimated lean mass from 65.0 to 62.0 kg.
Interpretation: The estimated split suggests lean tissue may be declining alongside fat. This is a prompt to review scan quality, food intake, strength training, hydration and functional measures. It is not direct proof of skeletal muscle loss or a basis for medication changes by itself.
Scale weight is steady while waist trends down
Baseline: 80 kg and 90 cm waist. Follow-up: 80 kg and 86 cm waist, captured under comparable conditions.
Interpretation: A lower waist with steady weight can still signal meaningful shape change. Review body-fat estimates, training adherence, strength and clinical context. Avoid claiming a precise amount of gained muscle without independent confirmation.
A large one-week improvement with almost no other change
Body-fat percentage falls sharply while scale weight and most circumferences hardly move. The client wore a loose shirt at baseline and fitted clothing at follow-up.
Interpretation: Scan conditions are not comparable. Do not present the apparent step-change as definite fat loss; establish a consistent new baseline and repeat.
A simple consultation sequence
- Validate: Was the new scan comparable to baseline?
- Interpret: Which measurements moved together and which contradict the trend?
- Act: Identify an appropriate coaching or clinical next step, and schedule the next comparable assessment.
Practitioner FAQ
Short, practical answers for staff and common client questions.
Is SNAP a diagnostic or medical imaging test?
No. The clinical partner kit presents SNAP as an assessment and monitoring tool. The scan does not directly image internal organs or diagnose a disease. Clinicians remain responsible for diagnostic and treatment decisions.
Is SNAP equivalent to DXA?
No. The underlying technology has been compared with DXA in studies, but a modeled estimate and a DXA measurement are not interchangeable for an individual patient. Use the appropriate method for the clinical question.
Why can SNAP differ from my gym scale or BIA system?
Methods use different physical inputs and estimation models. Clothing, hydration, timing, profile inputs and device-specific bias can all produce disagreement. For trends, follow one consistent measurement method and procedure.
Why did body-fat percentage increase while weight went down?
Possible explanations include relative changes in fat and lean mass, hydration, an outdated weight entry or inconsistent scanning conditions. Check the scan and inputs first; then seek confirmation with a subsequent scan or another method.
Can it show true muscle loss during GLP-1 treatment?
SNAP can provide modeled lean-mass and muscle-related indicators where available, but cannot directly quantify skeletal muscle tissue loss. Use strength, function, nutrition history and appropriate clinical methods when muscle preservation is important.
How often should clients be scanned?
Select a cadence aligned with the program and your ability to interpret meaningful change. Weekly or biweekly sessions may support engagement; monthly trends may be easier to interpret. Keep conditions consistent and do not overreact to small variations.
How should I handle a poor-quality or surprising scan?
Recheck clothing, hair, posture, rotation, lighting, profile details and weight. Repeat the scan when appropriate. If the result remains implausible, document it and rely on additional methods rather than forcing an interpretation.
Can clients wear leggings or a fitted shirt?
Tight leggings, compression shorts and a close-fitting top can be practical alternatives when privacy or comfort requires them. Avoid loose garments, and stay with the same clothing category for longitudinal comparisons.
Does waist on the report match my clinic’s manual tape protocol?
Not necessarily. Landmark definitions vary among devices and clinical tape protocols. Check the specific Fit3D landmark definition and consistently compare the same reported measurement over time.
Who should be assessed with additional caution?
Use extra caution for very lean or highly muscular bodies, pregnancy, significant edema, amputation and other unusual anatomy or situations not established in the published validation population.
Does SNAP store or transmit identifiable scan images?
SNAP uses Prism technology for body-scan processing. Fit3D’s privacy policy explains image and related data sharing with Prism; review the current privacy notice, consent process and your organization’s legal obligations before scanning.
Can I use SNAP results to select a drug dose or determine treatment eligibility?
Do not use a SNAP estimate as the sole basis for medical treatment, dosing or eligibility. Follow applicable clinical standards and use appropriate independent measurements and assessments.
For questions about privacy or consent, see Fit3D's current privacy policy.
Back to top ↑Evidence and references
Primary source material and boundaries for responsible scientific communication.
- Prism Labs (2026). A Consistent, Accurate Body Composition Alternative to DXA. A technical validation report that describes the study cohorts, reference methods, predictive-model variants, error summaries and repeatability. Read report.
- Prism Labs. Clinical Partner Kit. Technical explanation and clinical protocol source for the underlying Prism scanning technology. Used for methodological context, not to imply that every Prism application or feature is part of SNAP. Review reference.
- Prism Labs (2023). Announcement of independent evaluation conducted at Texas Tech University, including comparison with DXA and scan repeatability. This is a company summary of research; consult original publications when making formal statistical claims. Read summary.
- Fit3D. SNAP by Fit3D product information. Fit3D's published product positioning, tablet-based scanning workflow and client progress use cases. Visit Fit3D.
How to read validation numbers
| Term | Meaning | Common misunderstanding |
|---|---|---|
| MAE | Average absolute difference from a reference across observations. | Not a guaranteed maximum error for each person. |
| Correlation (r) | How closely two variables move together. | High r does not prove numerical agreement. |
| Bias | Average signed over- or underestimation. | Near-zero mean bias can hide large individual errors. |
| Repeatability | How consistent measurements are under similar conditions. | Highly repeatable readings can still have systematic bias. |
Recommended citation format in external materials: “Published research on the underlying Prism body-scanning technology reports [specified metric], under the study conditions described by Prism Labs (2026).” Include a direct link to the study.
Back to top ↑Fit3D Clinical & Wellness Partner Kit · Educational reference · Updated November 2026. Research figures attributed to their original publisher. This resource is not medical advice or a substitute for professional judgment.
Grow a healthier business
Get insights on how technology is transforming wellness businesses worldwide.