.png)
How Accurate Are Phone Body Scans Compared to DEXA?
Updated October 2026
Short answer: In published studies, phone body scans land about 2 to 4 body fat points away from a DEXA scan on average, and for one person the gap can be larger. Where phone scans shine is repeatability: two scans taken minutes apart usually agree within about 1 to 2 points, which makes them useful for tracking whether your body is changing over weeks and months. Use a phone scan for trends and a DEXA scan when you need a clinical reference number.
What DEXA is, and why it is the comparison point
DEXA (also written DXA, dual-energy X-ray absorptiometry) passes two low-dose X-ray beams through your body while you lie on a table, then separates bone, fat and lean tissue. It is the reference most body scan apps are tested against because it is widely available in research labs and clinics.
DEXA is not perfect either. Researchers treat a four-compartment model (which combines body water, bone mineral and body volume measurements) as a stronger reference. In one study of 141 girls and young women, DEXA body fat readings differed from the four-compartment model by 3.9 points on average, and individual readings could still be off by up to 6.7 points after correcting for that bias (Wong et al., 2002). Different DEXA machines can also shift results, which changes how well phone estimates appear to agree (International Journal of Obesity, 2026).
So when a phone app is "compared to DEXA", it is being compared to a very good tool, not to the absolute truth.
What the studies found
Study | Method tested | People | Average gap vs reference | Repeatability |
Tinsley et al., 2024 | Phone 3D scan (rotate in place, 3D avatar) | 131 adults | About 3.5 body fat points (mean absolute error) vs DEXA | Repeat scans within 0.5 to 0.7 points (technical error) |
Majmudar et al., 2022 | Phone photos analysed by AI | 134 adults | 2.16 points (mean absolute error) vs DEXA | Not the main focus |
Graybeal et al., 2022 | Phone photos turned into a 3D avatar | 102 adults | Equivalent at group level vs a 4-compartment model; underestimated body fat more as body fat rose | ICC 0.96 to 0.99 |
Wong et al., 2002 | DEXA itself | 141 girls and young women | 3.9 points average bias vs a 4-compartment model | Not tested |
Sources: Tinsley et al., Frontiers in Medicine, Majmudar et al., npj Digital Medicine, Graybeal et al., Clinical Nutrition, Wong et al..
A few details matter when you read these numbers:
Group averages look better than individual results. In the 2024 phone 3D study, the two estimation methods matched DEXA within a ±2 point equivalence margin for the group, but individual limits of agreement were about ±8.6 points (Tinsley et al.). Most people were close; some were not.
Many studies are funded by the company whose app is tested. The 2024 3D study was funded by Prism Labs, and the 2022 photo study was funded by Amazon (Tinsley et al., Majmudar et al.). Both were peer reviewed, but independent replication is still thin.
Lab conditions are strict. Participants in the 2024 study fasted for at least 8 hours, skipped exercise for 24 hours and wore minimal form-fitting clothing (Tinsley et al.). Your home scans will be more accurate if you copy those habits.
Results do not transfer between apps. Each app uses its own camera method and its own equations. A study on one app says little about another.
Accuracy vs repeatability: the part most people miss
Accuracy asks: is this number close to the true value? Repeatability asks: if nothing changed, would I get the same number again?
For tracking progress, repeatability matters more. If an app reads you 2 points higher than DEXA every single time, it still shows you, correctly, that you dropped 3 points over three months. In the 2024 study, the least significant change for the phone 3D scan was 1.4 to 1.8 body fat points (Tinsley et al.). That is the smallest change you can trust as real rather than scan-to-scan noise.
Worked example: is my change real?
Say you scan on the first Monday of each month, fasted, before training, in the same fitted clothes. Using the least significant change of about 1.8 points from the 2024 study as your threshold:
Month | Body fat reading | Change from month 1 | Bigger than 1.8 points? |
Month 1 | 31.0% | Baseline | |
Month 2 | 30.2% | -0.8 | No, could be noise |
Month 3 | 29.6% | -1.4 | No, but trending down |
Month 4 | 28.7% | -2.3 | Yes, likely a real change |
These readings are illustrative, not from a real user. The lesson: one month of small change proves little, but a steady direction across several scans, plus a total change past the threshold, is strong evidence. Pair it with your waist measurement and weight trend and the picture gets clearer still.
How to get the most accurate phone scan at home
Scan at the same time of day, ideally in the morning before eating or training.
Wear the same tight-fitting clothes and tie long hair up so it does not hide your neck and shoulders.
Use the same room, lighting and phone position each time.
Enter an accurate, current weight. Apps like Fitycal combine your scan with your weight to estimate fat and lean mass, so a wrong weight shifts the result.
Compare scans at least 3 to 4 weeks apart. Real body composition changes slowly.
When you should get a DEXA scan instead
A phone scan is a tracking tool, not a medical test. Book a DEXA scan if a doctor needs bone density, if you need a precise baseline for a study or competition, or if a medical decision depends on your body composition. Many people do both: one DEXA scan as an anchor, then regular phone scans to follow the trend.
FAQ
Are body scan apps accurate?
Reasonably, for most people. Published phone methods averaged about 2 to 4 body fat points from DEXA (Majmudar et al., Tinsley et al.), and they are very consistent from scan to scan, which makes them good for tracking change.
Is a phone body scan as good as a DEXA scan?
No. DEXA is more accurate for a single reading and also measures bone density. A phone scan is faster, cheaper and easy to repeat at home, so it is better suited to frequent check-ins.
Why does my body scan app give a different number than my DEXA?
Each method measures something different and uses its own equations, and even two DEXA machines can disagree. A gap of a few points is normal (Wong et al., International Journal of Obesity, 2026). Compare each tool only with itself over time.
How much body fat change is real on a phone scan?
In the 2024 3D phone study, changes smaller than about 1.4 to 1.8 points could be measurement noise (Tinsley et al.). Look for a steady direction over several scans.
Has Fitycal been validated against DEXA?
Fitycal has not yet published its own peer-reviewed validation study. We recommend using it the way the research supports: consistent conditions, monthly scans, and attention to the trend rather than any single number.
The bottom line
Phone body scans are not a replacement for DEXA, and honest apps should say so. They are, however, accurate enough on average and repeatable enough scan to scan to show you whether your training and eating are changing your body. If you want to see that change, try a 10-second scan with Fitycal on your phone, enter your weight, and compare your next scan side by side with your 3D avatar a month from now.
Comments