Deep learning-based automated hot-spot detection and tumor grading in human gastrointestinal neuroendocrine tumor

摘要

Ki-67 index is an important diagnostic factor in gastrointestinal neuroendocrine tumor (GI-NET). Thecurrent gold standard for grading GI-NETs involves the visual screening of histopathologically stainedtissues, for hot-spots containing high amounts of proliferating tumor cells (stained with Ki-67antibody). Subsequently, the Ki-67 index, i.e. the percentage of proliferating tumor cells within thehot-spot is manually obtained. To automate this subjective and time consuming process, we havedeveloped an integrated pipeline, termed SKIE (synaptophysin-Ki-67 index estimator), combiningdouble-immunohistochemical (IHC) staining for synaptophysin (stains tumor) and Ki-67, with wholeslide image (WSI) analysis. The Ki-67 index for 50 human GI-NET WSIs were estimated by SKIEand compared with three pathologists’ assessment, and the gold standard (exhaustive counting by afourth pathologist) based on the double-stained image. All four pathologists unanimously graded 38WSIs, among which, SKIE achieved 94.74% accuracy. One discrepant case was attributed to staininginconsistencies and the other to SKIE selecting a better hot-spot. The remaining 12 WSIs haddiscrepant grades among pathologists, and hence, the gold standard was chosen for comparison,wherein, 10 WSI grades matched with that of the gold standard, and SKIE assigned a lower and highergrade to two cases. Overall, SKIE agreed with the gold standard with a substantial linear weightedCohen’s kappa κ = 0.622 with CI [0.286, 0.958]. We further expanded our method to deep-SKIE,wherein, a deep convolutional neural network (DCNN) was trained and validated using 13,736hotspot-sized tiles from 40 WSIs, each categorized into one of four classes (background, non-tumor,tumor grade 1, tumor grade 2) by SKIE and tested on 9 WSIs. Deep-SKIE achieved an accuracy of91.63% with near-perfect agreement (κ = 0.88 with CI [0.87, 0.89]) with the gold standard.