The detection of fast particle-driven waves in the ion cyclotron frequency range (ion cyclotron emission or ICE) could provide a passive, non-invasive diagnostic of confined and escaping fast particles (fusion α-particles and beam ions) in ITER, and would be compatible with the high radiation environment of deuterium-tritium plasmas in that device. Recent experimental results from ASDEX Upgrade and DIII-D demonstrate the efficacy of ICE as a diagnostic of different fast ion species and of fast ion losses, while recent particle-in-cell (PIC) and hybrid simulations provide a more exact comparison with measured ICE spectra and open the prospect of exploiting ICE more fully as a fast ion diagnostic in future experiments. In particular the PIC/hybrid approach should soon make it possible to simulate the nonlinear physics of ICE in full toroidal geometry. Emission has been observed previously at a wide range of poloidal angles, so there is flexibility in the location of ICE detectors. Such a detector could be implemented in ITER by installing a small toroidally orientated loop near the plasma edge or by adding a detection capability to the ion cyclotron resonance heating (ICRH) antennae. In the latter case, the antenna could be used simultaneously to heat the plasma and detect ICE, provided that frequencies close to those of the ICRH source are strongly attenuated in the detection system using a suitable filter. Wavenumber information, providing additional constraints on the fast ion distribution exciting the emission, could be obtained by measuring ICE using a toroidally distributed array of detectors or different straps of the ICRH antenna.