Magnetic resonance imaging (MRI) can image brain structure non-invasively and without ionizing radiation. However, inferring histopathological information from MRI remains challenging due to lack of direct links between MRI signals and cellular structures. This study showed that deep convolutional neural networks can estimate histological staining intensity directly from MRI signals developed using multi-contrast MRI of the mouse brain. In this study, ex vivo MRI of mouse brain specimens were imaged with the skull intact and placed in a syringe filled with Fomblin to prevent tissue dehydration. The dataset contains histological data. This method provides three-dimensional maps of axons and myelin with tissue contrasts that closely mimic target histology and enhanced sensitivity and specificity compared to conventional MRI markers.