2019/07/22 by Muhammad Shaban, Shaban, Muhammad, Ruqayya Awan +9 · 3 citations
Computer Science · #AI in cancer detection #FOS: Computer and information sciences #FOS: Electrical engineering #Generative Adversarial Networks and Image Synthesis #Image Retrieval and Classification Techniques #Image and Video Processing (eess.IV) #Machine Learning (cs.LG) #Machine Learning (stat.ML) #electronic engineering #information engineering
paper · pdf · doi:10.48550/arxiv.1907.09478
openalex publication_date 2019/07/22 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Digital histology images are amenable to the application of convolutional\nneural network (CNN) for analysis due to the sheer size of pixel data present\nin them. CNNs are generally used for representation learning from small image\npatches (e.g. 224x224) extracted from digital histology images due to\ncomputational and memory constraints. However, this approach does not\nincorporate high-resolution contextual information in histology images. We\npropose a novel way to incorporate larger context by a context-aware neural\nnetwork based on images with a dimension of 1,792x1,792 pixels. The proposed\nframework first encodes the local representation of a histology image into high\ndimensional features then aggregates the features by considering their spatial\norganization to make a final prediction. The proposed method is evaluated for\ncolorectal cancer grading and breast cancer classification. A comprehensive\nanalysis of some variants of the proposed method is presented. Our method\noutperformed the traditional patch-based approaches, problem-specific methods,\nand existing context-based methods quantitatively by a margin of 3.61%. Code\nand dataset related information is available at this link:\nhttps://tia-lab.github.io/Context-Aware-CNN\n