How to convert RF absorption rate (SAR) values to electric field values? I'm currently reading a bunch of journal articles that describe biological effects of electromagnetic fields.  They often refer to the field in terms of specific absorption rate (SAR) values (W/kg).  I would like to figure out what electric field and/or magnetic field strength inside tissue that corresponds to (either RMS or peak to peak).
An example of such an article is
http://www.sciencedirect.com/science/article/pii/S0006291X15003988
I assume the dielectric properties of tissue would be required for this.  These are available for example from http://niremf.ifac.cnr.it/tissprop/htmlclie/htmlclie.php
The example article uses  0.04 W/kg SAR at 900MHz. Dielectric data for muscle at that frequency is:

To simplify things the field strength can be assumed to be uniform throughout.
 A: If use the example above, by relative permittivity and loss tangent, one can get the imaginary part of the dielectric constant, 
see below:
https://en.wikipedia.org/wiki/Dielectric_loss
once know the permittivity, the refractive index can be easily caculated, and if the material is not a gain media, then imaginary part for the refractive index is always positive. you can double check the data with the penetration depth. see:
https://en.wikipedia.org/wiki/Penetration_depth
E=E0*e^(-z/(2*l)), where l is the penetration depth. you will get the E field for  depth z if you know the incidence E field(E0). For E field on other direciton, your information is not enough. 
Specific absorption rate (SAR) in your question is most likely to be one average value( otherwise it has to be a funtion of a lot of paramters, such as position, organ et.al).Generally it's a volum integration,see: 
https://en.wikipedia.org/wiki/Specific_absorption_rate
If we assume the issue is not magnetic , to simplify
at depth z,
Absorbed Energy=conductivityE^2=conductivityE0^2*e^(-z/l)
Total absorbed energy would be the integration for above with depth.
You have to know another parameter(the thickness of the sample,h) to do the integration, finally:
The total energy=conductivityE0^2(hl)(1-e^(h/l))
And the total power is the time average for the total energy,if E0 is a sine wave of time,then total power would be the half of above.
And you also need to know the total mass(M) for your sample, now:
SAR=the total power/M=1/2*conductivityE0^2(hl)(1-e^(h/l))/M
Now we can retrieve the initial E field(E0), and can get the E field for depth z. The general process would be like this, when more parameter were provided, you may be able to modify it to a more accurate version. 
