The data is divided into two distinct datasets: one dedicated to mice undergoing remyelination, known as the validation dataset, and the other focusing on a healthy control specimen. These datasets consist of transmission electron microscopy (TEM) images of the corpus callosum (CC) in adult mice. Both datasets are enriched with annotated ground truth information for two key tasks: instance segmentation (identifying individual myelinated axons) and semantic segmentation (discerning axons, inner cytoplasmic tongue, and myelin). Furthermore, we have incorporated ilastik pixel and object classifiers, specifically trained on remyelinating data, into the repository. To streamline usage, the training dataset has been integrated into the corresponding project file.