Machine Learning Cancer Diagnosis Based on Medical Image Size and Modalities

Abstract

Nowadays, machine learning (ML) is one of the significant technology in the new era. It almost applies to all the fields in our lives. ML has a notable impact on medical care when dealing with medical images. Medical images are used to investigate the internal-parts of the body of human. The type and size of medical images vary from one scan to another. Besides, medical images are not like natural images. While natural images can perform object detection and classification easily, even if the images are re-sized to smaller images. Resizing medical images is not an efficient method because they interact with pixel-level. In this chapter, medical imaging modalities and histopathology are explained. Furthermore, the best medical image type and size for classification and detection of medical diagnoses are explained. Moreover, specific methods are considered in medical images such as image compression, image format, image resize, and other essential aspects. Finally, we also give a brief summary of deep learning algorithms that are used with medical images.