This study analyzes urban verticalization and socio- spatial patterns in Belo Horizonte using LiDAR-derived building heights and infrastructure indicators. A normalized Digital Sur- face Model (nDSM) was generated from LiDAR data to map vertical structures with high spatial accuracy. Self-Organizing Maps (SOM) were applied to cluster neighborhoods based on height, infrastructure, and demographic data. The results reveal distinct urban typologies, including central consolidated zones, peripheral vulnerable areas, and hybrid transitional regions. While verticalization reflects urban consolidation, it must be interpreted alongside socio-infrastructural conditions to fully understand spatial inequalities and urban dynamics.