Artificial Intelligence and Cancer Detection

Abstract

Artificial Intelligence (AI) has recently been used to support medical diagnosis, developing tools to increase diagnostic efficiency and decrease response times, particularly in cancer detection with digital images. Algorithms developed based on AI are accurate; however, external or independent validation is needed to ensure that these algorithms are generalizable. This chapter describes the steps to build models based on AI, which include data acquisition (with clinical data acquired by electronic health records or provided by the patient himself and clinical images by instruments such as X­Rays, CT scans, and MRIs, among others). Next, some techniques for pre­ processing the acquired data are shown to ensure the data quality. Afterward, some methods for data processing are shown, including feature extraction, classification, and segmentation methods. Some recommendations are also demonstrated for visualizing and presenting the findings in medical images. Finally, the performance metrics must be considered to evaluate the model or algorithms developed, such as recall, accuracy, dice score, sensitivity, and so far.