Team: Gordon Center for Medical Imaging, Massachusetts General Hospital, Harvard Medical School, USA
Authors: Aoxiao Zhong, Quanzheng Li
This is a reproduced version of the algorithm by the top-performing team in the challenge event (HMS-MIT Method 1). I replaced the GoogLeNet they used with a deeper CNN model ResNet-101. The model is trained based on a pre-trained ResNet-101 for ImageNet classification. The classification task is based on features extracted from tumor probability map with a random forest classifier as described in the submission by HMS-MIT Method 1. Further postprocessing step was used to remove candidates that are too small for the lesion-detection task.
The following figure shows the receiver operating characteristic (ROC) curve of the method.
The following figure shows the free-response receiver operating characteristic (FROC) curve of the method.
The table below presents the average sensitivity of the developed system at 6 predefined false positive rates: 1/4, 1/2, 1, 2, 4, and 8 FPs per whole slide image.
Error rendering graph from file