Speaker
Description
Systems using machine learning can learn from initial data labeled, then identify features and make decisions with little human intervention. An input image can be converted into millions of pixels for classification tasks. Accordingly, data entry will make processing very difficult. To make it easier, the image is converted into a reduced set of features. This selects and measures representative properties of the raw input data in reduced form and size. Furthermore, a streamlined set represents the relevant piece of information needed to perform the desired task. This set can be represented by color, texture, shape, or a simple part of an image. In this essay, some studies on the use of machine learning to process medical images in the diagnosis of several diseases such as ophthalmology, dermatology, osteoarthritis, and hematology will be summarized.
