I am wondering how the hounsfield values are handled when visualizing CT datasets. Since the range of the HU unit contains negative values which might be problematic to handle (for example when multiplying with an attenuation factor during ray casting), I am assuming they are converted to the range [0,1], where 0 represents air (-1000 HU) and 1 bones (~3000HU).
Am I correct in this assumption? How is the problem that not every real number can be represented by an floating point dealt with?