A three-dimensional Direct Numerical Simulation (DNS) database of statistically spherical flames propagating into the mono-sized droplet-laden mixtures for an overall (i.e. gaseous + liquid phases) equivalence ratio of unity has been considered to investigate the distribution of flame topologies associated with flame self-interaction events. The analysis has been carried out for different values of initial droplet diameters and turbulence intensities to demonstrate their influences on the statistics of flame self-interactions. Spherically expanding stoichiometric premixed gaseous flames have also been additionally considered for the same initial burned gas radius and turbulence intensities for the purpose of comparison with the cases propagating into droplet-laden mixtures. It has been found that the topologies associated with flame self-interaction predominantly occur close to burned gas side of the flame front for turbulent premixed flames and also for droplet cases with small turbulence intensities, although tunnel closure/formation topologies are prevalent in premixed flames in contrast to unburned mixture pocket type topologies in the droplet cases. The distribution of the flame topologies has been observed to be influenced by both turbulence intensity and droplet diameter. The flame self-interaction events have mostly been found in the preheat zone for highly turbulent spray flames. It has been found that the occurrence of tunnel formation and closure topologies dominates over the availability of the unburned and burned mixture pockets in all cases except for the droplet cases under low turbulence intensities. The frequency of flame self-interaction events increases with increasing turbulence intensity and droplet diameter. The differences in the nature and distributions of flame self-interaction events between turbulent premixed and droplet cases are likely to have important implications on the possible extension of flame surface based modelling methodologies for simulating turbulent combustion in droplet-laden mixtures.