Breast Cancer Detection Using Artificial Intelligence Approaches

Abstract

The rising rate of breast cancer is a global health crisis that affects women of all socioeconomic backgrounds for several reasons. As a result, effective screening is crucial for early diagnosis and treatment. There are several ways in which artificial intelligence (AI) is already changing our lives for the better. Adding AI to the current screening process streamlines and simplifies the whole operation. Benefits of using AI approaches in breast cancer screening include faster and more accurate outcomes. However, there are several obstacles along the path to AI integration that must be methodically addressed. The majority of inherited disorders are caused by changes in a single gene. To better clarify how genes contribute to illnesses with a complicated pattern of inheritance, such as diabetes, asthma, cancer, and mental illness, is one of the most challenging tasks ahead.There is no one gene that can definitively determine whether a person will become sick or not. Several genes may each subtly contribute to a person's vulnerability to a disease; genes may also alter how a person responds to environmental variables. It is probable that more than one mutation is necessary before the illness is visible. Machine learning is a subfield of artificial intelligence that uses a wide range of statistical, probabilistic, and optimization approaches to help computers learn from their experiences and see hidden patterns in otherwise difficult-to-interpret data. Medical uses, especially those that rely on complicated proteome and genomic measurements, may benefit greatly from this capacity. Among females, breast cancer is high as a major killer. Millions of women throughout the world may benefit greatly from an improvement in their quality of life if breast cancer were detected earlier. Since automating early detection and diagnosis is so crucial, many Convolutional neural networks (CNNs) and other deep learning models might possibly improve detection accuracy by learning complicated characteristics directly from the photos themselves. Researchers have used mammograms, ultrasonography, magnetic resonance imaging, histopathological pictures, and any combination thereof to automate the process of breast cancer identification. This dissertation compares and contrasts the benefits and drawbacks of each of these imaging techniques. It also includes a directory of places where the datasets may be accessible for study.