Radiomics applied to pulmonary infection: A review

Abstract

Pulmonary infective diseases with high morbidity and mortality have contributed significantly to the socioeconomic burden worldwide. Despite advances in the technology for identification of pathogenic microorganisms, increased diagnostic accuracy for noninfectious pulmonary inflammation, and the availability of effectively individualized therapy, a major concern is the lack of optimal diagnostic approaches and biomarkers for the early stages of infectious disease. Radiomics, a concept first proposed in 2012, is a process that can extract quantitative features and high-dimensional data from digital medical images such as conventional chest X-ray, computed tomography (CT), magnetic resonance imaging, positron emission tomography, and single-photon emission CT. The high-dimensional data and mineable features can be utilized to build descriptive and predictive models to quantify the inflammatory burden on the lung and understand lung function. Overall, radiomics now plays an important role in the clinical diagnosis and treatment of pulmonary infection, with medical imaging being routinely applied in the clinical workup of pneumonia. Radiomics may also provide an opportunity to allow accurate diagnosis of pulmonary infection in the early stages, thereby decreasing mortality at a low cost.