A Review of AI Devices in Cancer Radiology for Breast and Lung Imaging and Diagnosis
Shivansh Khanna
School of Information Sciences, University of Illinois at Urbana-Champaign
Keywords: artificial intelligence, breast imaging, deep learning, diagnostics, digital breast tomosynthesis, lung imaging, mammograms, medical imaging
Abstract
The AI integration in oncology has transitioned from a speculative concept to a concrete reality. Across the stages of cancer care, various AI applications are at different stages of development and are being incorporated into practices. These range from early detection and diagnosis to treatment planning and follow-up care. This research outlines an inventory of devices introduced between 2015 and 2020 in three categories: devices for breast imaging and analysis, lung Imaging and Analysis, and devices for image reconstruction, quality assessment, and other diagnostic support. The Breast Imaging and Analysis category primarily features devices that target breast-related images, particularly in breast cancer context. These tools assist in detecting suspicious regions in mammograms, categorizing breast tissue types, and elevating the precision of breast cancer diagnoses. Such technologies are mostly compatible with mammograms and digital breast tomosynthesis (DBT) exams. In the Lung Imaging and Analysis segment, the devices listed are mainly developed to aid healthcare professionals in identifying, documenting, and analyzing pulmonary anomalies. Given the global significance and fatality rates associated with lung cancer, these devices aim to enhance the precision of detecting nodules, lesions, and other potential signs of lung ailments, primarily through CT chest exams. The image reconstruction, quality assessment, and other diagnostic support category encompasses devices focusing on general medical imaging. These devices employ innovative algorithms and deep learning methodologies for tasks such as image reconstruction, quality evaluation, and specific condition detection, including prostate cancer and colorectal polyps. The category also covers tools designed for radiation therapy planning.
