A rigorous mathematical construction of Feynman path integrals for the Schrödinger equation with magnetic field (Q2191179)

As is known, the path integral formalism used extensively in quantum field theory is not well defined in the strict mathematical sense. While this formalism is very instructive for calculating expectation values in quantum field theory, the rigorous mathematical definition of path integrals is a quite complicated matter and involves a couple of quite abstract infinite dimensional spaces of which the Hilbert space is best known in physics. Still, as experts in this field, the authors find their way of argumentation in this area. The main aim of this publication is to show that under these mathematical abstractions, the path integral is indeed bounded and can be defined properly, even for the application it is usually intended for in physics, namely a perturbation series expansion. As an example, the authors consider the solution of the Schrödinger equation under the action of a vector potential. By employing the slicing method, they demonstrate that a consistent definition of the path integral does not lead to the (simpler) Itô integral, where the integrand is taken at one of the boundaries of the slice, but instead to the (more involved) Stratonovich integral centered in the middle of the slice. Rather, they can show that their treatment leads unavoidably just to the Stratonovich integral. This publication is helpful for those who are in doubt whether the path integral formalism can be defined properly and are eager to get to know more about such methods. Mirrored back to physics, the results can give a hint at which provisions for the usual slicing method are most appropriate in mathematical terms.