It will, however use of a blade is not feasible for a number of reasons.
1) The human hand is not precise enough to take cross sections.
2) The human breast would deform under physical cutting, which would ruin data.
3) The same breast would need to be sliced multiple times at different angles.
Instead, what would need to be done is use cameras and motion sensing equipment to generate a 3D model of the human breast. The wire-frame stuff they use in CG animation would have to be avoided, because we need to maintain the actual surface curvature. So when I say motion sensor, it should probably be read as location sensor. We want to have the shape captured by a computer with at least this fidelity and resolution:
Now, we could make it solid but we're really only concerned with the boundary surface. Making the object hollow conserves a lot of memory. Crop the breast, throw it into a computer algebra system, and use an interpolating function to generate a parametric surface. Then, define various planes and check their intersection with the boundary of the breast. That would give you a space curve to compare to various conic sections. For grins, we could put both breasts side by side, pass the plane straight out of the woman's body and see if they generate hyperbolas.
For rigor's sake, we'd probably want to get around twenty-five women evenly spread across various cup sizes. Obviously, they need to be natural breasts without any surgical procedures to modify shape or size. Breast implants, reductions, or lifts all involve the use of computer and machine guidance to artificially impose a geometric structure on the breast. That contaminates the data.