The article from which the x-ray images were taken went to some length to say the 3D interpretation was inaccurate, and they had a holographic 3D x-ray process that produced non-arbitrary 3D rendering.
In case you forgot, re-creating the 3-D information from even a non-holographic X-ray is not 'arbitrary', even though it might be inaccurate.
There are real human bodies (maybe even of the person X-rayed) which we can use to 'tune' the parameters in the software: we all know that the coccyx is not longer than the femur. ;-)
That being said, using multiple light sources does give more accurate 3-D information than a single source of light. But anyone who knows of scattering theory or of partial derivatives would be able to point that out.
Cheers!