Application value of CT diagnosis technology based on Myrian image post-processing system in screening and diagnosis of early lung cancer

  Background  The incidence of lung cancer is increasing year by year, and it is of great importance for lung cancer patients to receive early diagnosis and treatment, to improve the 5-year survival rate, and avoid some benign pulmonary nodules from receiving unnecessary surgical intervention.

  Objective  To evaluate the CT diagnosis technology based on Myrian image post-processing system in early lung cancer screening and diagnosis.

  Methods  From November 2016 to May 2019, 70 patients with pulmonary nodules who were suspected to be early lung cancer were selected as the study subjects. All patients received chest spiral CT examination, and confirmed the nature of the nodules by pathology results after surgical resection. The original CT thin-layer axial image data were sent to the workstation, and the three-dimensional reconstruction of the lesions, surrounding lung tissue and trachea was performed by Myrian post-processing software.

  Results  There were 94 nodules in 70 patients, which were confirmed by pathology as 56 malignant and 38 benign nodules. CT scan showed that all the nodules were less than 3 cm in diameter. Myrian postprocessing diagnosed 57 nodules as malignant and 37 as benign. Compared with the gold standard, the accuracy of Myrian postprocessing in diagnosing early lung cancer was 88.30%, with the sensitivity and specificity of 91.09% and 84.21% respectively. Univariate logistic regression analysis showed that shape, lesion location, nodule volume and border shape obtained after Myrian reconstruction were of predictive value for malignant nodules (P＜0.05). Multiple logistic regression analysis showed that the size of nodules ≥ 500 mm³, and irregular, lobulated or burr-like boundaries, had the value of predicting malignant nodules (HR＜1, P＜0.05).

  Conclusion  Common CT images can be reconstructed into three-dimensional images by Myrian post-processing technology, which can clearly show the spatial relationship between tumor and surrounding trachea, aiding the diagnosis of early lung cancer with high accuracy.