The nuclear lamina functions as a structural support system for the nucleus and must be broken down during mitosis. Not much is known about the process by which this breakdown is initiated. Prior studies have suggested that the nuclear envelope is depleted of lamin B1 at high curvatures, which destabilized the nuclear lamina and possibly plays a role in mitotic nuclear envelope breakdown. A statistical probability model has been proposed to describe this depletion based on the Gaussian curvature of the nuclear lamina. However, this model has only been tested on limited two-dimensional data. In this thesis, a quantitative image analysis pipeline is created and applied to test the relation between the absolute Gaussian curvature of the nuclear lamina and the density of lamin B1 in three-dimensional confocal images. To accomplish this, the nuclear lamina are segmented with unsharp-masking and steerable filters. The curvature of the nuclear lamina is computed using a grey-scale surface curvature estimation. Two different density states of the lamin B1 meshwork are found and a transition from high to low density is shown for increasing curvatures. These results support the previously proposed probability model for the depletion of lamin B1 at high curvatures and provides a deeper understanding of the structure of the nuclear lamina. Additionally, the image analysis pipeline established in this thesis provides helpful tools for effective segmentation and curvature estimation of the nuclear lamina and can be applied in future research.