Performance measures are an important tool for assessing and comparing different medical image segmentation algorithms. Unfortunately, the current measures have their weaknesses when it comes to assessing certain edge cases. These limitations arouse when images with a very small region of interest or without a region of interest at all are assessed. As a solution for these limitations, we propose a new medical image segmentation metric: MISm. To evaluate MISm, the popular metrics in the medical image segmentation and MISm were compared using images of magnet resonance tomography from several scenarios. In order to allow application in the community and reproducibility of experimental results, we included MISm in the publicly available evaluation framework MISeval: https://github.com/frankkramer-lab/miseval/tree/master/miseval