Associação entre os scores radiômicos e fleischner para probabilidade de malignidade em nódulos pulmonares incidentais

Abstract

Pulmonary nodule is a rounded opacity in the lung, at least partially delimited, smaller than 3 centimer with soft or calcic density. The existence of a the solitary pulmonary nodule, detected in incidental way or during a routine examination of the image has become the tool point for the early detection of lung cancer and its correct interpretation is a still scientific challenge. The main manifestation of lung cancer in an initial phase is pulmonary nodule, however the errors of interpretation of images make its classification in malignant or benign an extremely difficult task. Is currently used for the screening of lung cancer by the discovery of a pulmonary nodule the recommendations of an important study: Fleischner Society Pulmonary nodule recommendations but there are a lot of limitations for this. Radiomics is a computational image study that intends to extract the maximum of characteristics of the image to assist in the support of decision using mathematical models. The objective of this study was to verify the association between radiomcs data from pulmonary nodules and the Fleischner scores in order to improve the diagnostic accuracy of malignancy prediction of a randomly detected pulmonary nodule in a chest computed tomography. We studied 210 radiomcs attributes extracted from chest computed tomography and after sought an association with the Fleischner data in order to strengthen the radiomcs data to possible future Validations. Of the attributes studied, we found a correlation between 7 of them, in this case providing to improve the medical decision and influencing the survival of lung cancer patients, which diagnosis and the effective treatment process is a scientific challenge.