Computational studies can be used to support the development of peripheral nerve interfaces, but currently use simplified models of nerve anatomy that do not accurately represent anatomical conditions. To better quantify and model neural anatomy across the population, we have developed an algorithm to automatically reconstruct accurate peripheral nerve models from histological cross- sections. We acquired serial median nerve cross-sections from human cadaveric samples. We developed a processing pipeline involving four steps: registration, detection, segmentation, and reconstruction. While our algorithm could not always outperform simplified anatomical models, it nonetheless provided useful anatomical information that would otherwise be lost. Fascicle detection by convolutional neural network worked particularly well and can be applied to histological images as is. The other components of the pipeline need improvements in order to build truly accurate models. This work provides a baseline from which to progress toward a fully automatic approach to constructing peripheral nerve models.