Cutting-edge AI with third-party validation

Body fat percentage is one of the top predictors of long-term health. Yet access to accurate body composition results has been expensive and time consuming, or the more affordable solutions have been inconsistent and inaccurate.

Prism Labs has made affordable, accurate body scanning more accessible through advancements in computer vision technology, producing results in line with DXA, the gold standard for body composition tracking.

Why this is a breakthrough

The Prism Bodymapping solution introduces three groundbreaking technology innovations:

  1. Unlike other 3D-body scanning solutions that require a rigid subject, the person being scanned can be moving. This enables them to turn themselves around in front of the camera without a turntable for a simple user experience
  2. Unlike other 3D-body scanning solutions that require depth cameras with LIDAR capabilities, the Prism solution works with a color camera which reduces hardware costs
  3. The Prism scanning solution is integrated with the iPhone’s advanced camera capabilities for skeletal tracking, improving the user experience and accuracy as well as enabling integration by fitness and health solutions like Tempo

How it works

Prism is the only true 3D body reconstruction via iPhone solution in the market that provides accurate body metrics. Other solutions rely on statistical body models, not a true body reconstruction that looks just like you.

First, the single camera captures you as you spin in front of it. These images are then used to generate a 3D reconstruction. Once we've created your 3D avatar, our machine learning algorithms perform inferences on your avatar which provides you with a highly reliable body composition result.

Third-party validated

The only iPhone bodymapping technology to be third-party validated.

Prism Labs works closely with Dr Grant Tinsley, a leading expert in body composition from Texas Tech University, to validate its methods and results against the most accurate standards in the field.

Our most recent results have found that our results are highly consistent, meaning that when you do back-to-back scans under the same conditions you will get the same results. Our error rate for body fat percentage is ± 0.20 % and for body circumferences is ± 0.45%.

Additionally, we found that our absolute median error rate for accuracy is ± 3.9% for body fat percentage when compared to DXA results, and ± 2.2% relative median error for body circumferences when compared to measurements taken with a depth camera-based 3D body scanner.

Read about the latest peer-reviewed paper validating our single camera system.