for a rough estimation please use the KERMA (kinetic energy released in material) as energy dose. KERMA is an upper estimation of dose; but for sufficiently large homogenous sample sections (a few cm in diameter) the KERMA is equal to the dose at the center of these sections.
Please have a look at the attachment, taken from the famous book of Frank. H. Attix (Introduction to radiological physics and radiation dosimetry).
Please use the µen coefficients for the calculation.
Almost everything has been answered by Martens sir. You may initially see absorbed dose than radiation weighting and tissue weighting factors leading to effective dose. Note that beta radiations will deliver only skin dose where as photon from Cs or Co will also deliver dose to deeper organs. Individual doses to organs can also be calculated. Finally one has to find energy deposited per unit mass. Advance techniques are montecarlo based simulations.
Absorbed dose in a medium can also be calculated from the quotient of radiation energy deposited per unit mass of the medium : D (Gray) = dE/dm.
If you measure radiation exposure in air, you can convert it to absorbed dose by means of a conversion factor with respect to the medium of interest.
You can also measure it by using ionization chamber, Thermoluminescent dosimeter (TLD) and film. By exposing these detectors (prior calibration) to ionizing radiation, you can determine the absorbed dose in each detector. You can use a phantom (water or plastic) to represent human body.
If you measure radiation exposure in air using calibrated ionization chamber connected with electrometer, you get air Kerma, and you convert it to absorbed dose by means of a conversion factor with respect to material and energy of radiation.
It should be considered if the source is external or distributed internally inside the patient, like in nuclear medicine applications (diagnostic or therapy). In the last case the distribution of radiopharmaceutical in the organs or total body must be determined by measurements or published data could be also used sometimes. The MIRD formulation or the ICRP methods and biokinetic models are used for estimation of absorbed dose in patient or occupational-workers. Medical applications for imaging using X rays, like CT have particular magnitudes and formulation for estimation of absorbed dose in patients.
Now a days you may directly model in codes say monte carlo or so on. Just model the geometry and all equations are in built. Just select appropriate parameters which are meaningful and are demanded by softwares. Many ICRP/ICRU works are more or less just monte carlo studies with reasonable judgment and few experimental proofs.
Great information has been written in the comments of professional people in the field. Moreover, absorbed dose can be measured thermoluminescent dosimeter (TLD).