SKIE: An Automated Approach to Quantitation of Ki-67 Index from Human Gastrointestinal Neuroendocrine Tumor

BACKGROUND: Ki-67 index is an established diagnostic factor in gastrointestinal neuroendocrine tumors (GI-NETs). Automated assessment of this index is challenging due to the difficulty in computationally distinguishing tumor from non-tumor regions in standard histopathological images. We propose an integrated approach, termed SKIE (Synaptophyin-Ki-67 index estimator), combining whole slide image (WSI) analysis with double immunohistochemical staining for Ki-67 and synaptophysin (stains tumor regions).