Lung Nodule: Imaging Features and Evaluation in the Age of Machine Learning

AbstractPurpose of ReviewWith the unprecedented increase in chest CT studies, especially due to implementation of lung cancer screening, evaluation of lung nodules by radiologists can be exhausting and time-consuming. Machine learning promises to be a useful tool for detection and characterization of nodules. The purpose of this review is to evaluate the recent literature pertaining to machine learning in lung nodule detection and evaluation.Recent FindingsThere has been a recent surge of publications pertaining to machine learning and its applications in chest imaging. Many studies have shown promising results for automatic detection and characterization of lung nodules. Other studies have shown combined performance of a radiologist and computer-assisted detection (CAD) out performed a single radiologist, CAD alone, and double readers. Although these recent advances heighten expectations, it is important for developers and users to be mindful of challenges such as training, validation, independent testing, and proper user training.SummaryComputer-aided technology can help radiologists in evaluating lung nodules especially with the large number of scans performed. Recent advances in machine learning are replacing traditional methods and could significantly change the way radiology is practiced.

Source: Current Respiratory Care Reports - July 21, 2019 Category: Respiratory Medicine Source Type: research