In this work, we used a database consisting of 50 patients diagnosed with Chronic Migraine (CM) or Episodic Migraine (EM), based on the occurrence of headache with a frequency equal to or greater than 15 days per month for more than 3 months (CM) or fewer than 15 days (EM). These patients underwent two clinical studies: a Neuropsychological assessment (NPA) and Magnetic Resonance Imaging (MRI). Numerical variables were obtained from both studies, forming a tabular database. We applied three clustering models (K-Means, Mean Shift, VBGMM), obtaining two robust groups that were consistent across all three models. We found that there are key features in these groupings corresponding to both psychological descriptors (Anxiety in NPA) and anatomical descriptors (total gray matter concentration and in certain cerebral Brodmann areas). Additionally, each of these two clusters contained patients diagnosed with both CM and EM. We conclude that a more detailed characterization is necessary in the diagnosis of the pathology and that the diagnoses of Chronic Migraine (CM) or Episodic Migraine (EM) may be insufficient when addressing treatments.