Methods and Systems for Human Tissue Analysis using Shearlet Transforms

摘要

Various arrangements for identifying and grading cancer in tissue samples are presented. A digital image of a stained tissue sample may be acquired. A Shearlet transform may be performed on the digital image of the stained tissue sample. Shearlet coefficients may be calculated based on the performed Shearlet transform of the normalized digital RGB image of the stained tissue sample. A trained neural network may be applied to create a plurality of feature maps using the digital image and Shearlet coefficients, wherein the trained neural network was trained using a plurality of images and Shearlet coefficients of a plurality of digital images. A classifier may be applied to an output of the trained neural network to identify whether cancer is present in the stained tissue sample. A notification may be output that is indicative of a grade of detected cancer in the sample.

主权项

1. A method for identifying and grading cancer in tissue samples, the method comprising:
acquiring, by tissue sample processing hardware from a medical tissue scanner, a digital RGB (red, green, blue) image of a stained tissue sample, the stained tissue sample to be analyzed on a multiple level scale for cancer, wherein the stained tissue sample is of a type of tissue; normalizing the digital RGB image of the stained tissue sample to correct for medical tissue scanning and staining variations; performing a Shearlet transform on the normalized digital RGB image of the stained tissue sample; calculating Shearlet coefficients based on the performed Shearlet transform of the normalized digital RGB image of the stained tissue sample; following the Shearlet transform being created and the Shearlet coefficients being calculated, applying a trained neural network to create a plurality of feature maps using the Shearlet coefficients, wherein the trained neural network was trained using Shearlet coefficients of a plurality of digital RGB images, the plurality of digital RGB images comprising a first subset of digital RGB images that are indicative of tissue of the type of tissue having varying grades of cancer and a second subset of digital RGB images that are indicative of non-cancerous tissue; applying a classifier to an output of the trained neural network to identify whether cancer is present in the stained tissue sample; and outputting a notification indicative of whether cancer is present in the stained tissue sample based on applying the classifier to the output of the trained neural network.