4. Diagnosis of Breast Cancer by Malignant Changes in Buccal Epithelium Using Artificial Intelligence, Internet of Things, and Cloud Storage

The aim of the paper is to describe a project of novel highly accurate AI system for screening of breast cancer based on investigation of fractal properties of chromatin in Feulgen-stained nuclei of buccal epithelium. The work of the system consists in data collection in cloud storage, morphological pre-processing and classification on the cloud platform and traditional computers, and communication with patients using IoT technology. We studied data on 130 patients: 68 patients with breast cancer, 33 patients with fibroadenomatosis, and 29 healthy individuals containing in average data on 52 nuclei of buccal epithelium. The data set consists of 20,256 images of interphase nuclei of buccal epithelium (6752 nuclei scanned without filter, through a yellow filter, and through a violet filter). Each image consists of three channels: red, green, blue, as well as gray halftones. The combination of cloud storage of data, classification on cloud platform and using the IoT technologies for communication with patients provides high accuracy of diagnosis and comfort for patients.