The PHANGS-MUSE survey offers us an unprecedented view of the ionized interstellar medium of nearby star-forming galaxies. The high spatial resolution (ca.50 pc) allows us to identify and isolate single emitting clouds, while the broad spectral range offers us the opportunity to study their properties in detail. However, a correct classification of the nebulae (e.g., HII regions, planetary nebulae, supernova remnants) is critical to studying them and their role in galaxy evolution. Many classification schemes have been developed in the literature in the last few decades. Still, none of them can take advantage of all the information provided by modern integral field spectrographs like MUSE at the VLT. In this talk, I will present the results of a new algorithm that aims to reliably and objectively classify nebulae using the Bayesian principle of the odds ratio to compare spectral and morphological properties of each region to models representing different classes of nebulae. This algorithm is applied to the nebulae identified in the PHANGS-MUSE survey to produce the most extensive catalog of classified ionized nebulae available in the literature.