you know that the x-ray attenuation is dependent on a) the atomic number (increases with Z) and b) on the x-ray photon energy E in use (decreases with increasing E).
Soft tissues (as well as blood and tumor tissue) mainly consist of low Z elements. So you have to go for low energies (i.e. low tube voltage kVp) to have the best image contrast.
However, unfortunately low voltages imply low tube currents. So for extreme low voltages your images will suffer from statitical noise of the acquired data. So you have to agree a compromise between the current, acquisition time and low voltage.
You will come out with tube voltages of around 50 to 70kV.
When doing µ-CT at a synchrotron beamline there is no such stringent (current) restriction. Here you may go down to around 20keV or even lower.