Preprocessing of Whole-Slide Images Breast Cancer to analysis metastasis Lymph Node using Machine Learning

Practical Experimentation

Abstract: Breast cancer is the most common invasive cancer and the second leading cause of cancer death in women and regrettably. The prognosis for breast cancer patients significantly deteriorates when breast cancer metastasizes are identify. The gold standard for breast cancer diagnosis currently is the histological analysis of a suspected section. Whole slide imaging (WSI), which refers to scanning conventional glass slides in order to produce digital slides \cite{wsiinpathology}, is the most recent imaging modality being employed in the pathology environment. The emergence of whole-slide imaging in digital pathology strengthened the advancement of examination of suspected sections via digital image analysis with deep learning techniques. The WSI images can be analyzed via a variety of algorithms. However, identify cancerous regions in Whole Slide Imaging (WSI) using Deep Learning (DL) with efficiency can be a challenge for different perspectives. One of those is that WSI is images with billions of pixels, variations in coloring with differences in tissue thickness. In this way, Preprocessing those images is necessary to reduce problems that can be faced in the automatic analysis of WSI. This study reviews the techniques that can be applied to preprocessing WSI to mitigate issues considering a deep learning analysis perspective..