Quantifications of COVID-19 radiological pulmonary structures (Coronavirus infection)

In December 2019, a series of cases were identified in China of the so-called COVID2019, which has spread throughout the world, currently reaching more than 2,000,000 cases with 140,000 deaths. Currently, viral nucleic acid analysis based on RT-PCR (real-time Polymerase Chain Reaction) is used as a standard reference method to confirm COVID-19 infection. One of the main consequences is a pulmonary inflammatory process, which leads to a large part of deaths and hospitalizations. In this sense, computed tomography (CT) is the method of choice for these analyzes and has been considered the main imaging method with high sensitivity for detecting the consequences of COVID19 and even considered as a diagnostic method. COVID19 causes several lung damage, mainly inflammatory processes that manifest themselves with ground-glass radiological findings, among others. The evaluation is subjective and requires trained specialists. Improvements in visualization and quantification of structures and radiological findings are essential for a better diagnosis and quantitative determination of the affected pulmonary percentage. In this proposal, using image processing, generate algorithms to enhance the affected lung characteristics and quantify these structures in relation to the lung volume. Methods of filtering, segmentation and extraction of objective characteristics in CT images will be used in the initial phases, during and after negative by RT-PCR, aiming at a temporal and spatial characterization, in front of the different pulmonary areas. This type of objective analysis will provide greater support for the diagnostic decisions of radiologists and clinicians, providing data support for the treatment and sequencing of cured patients and generating technological processes to aid diagnosis and support classificatory data involving pulmonary structures affected by COVID19. (AU)