Diagnostic Value of Deep Learning-Based CT Feature for Severe Pulmonary Infection.

Tinglong Huang,Xuelan Zheng,Lisui He,Zhiliang Chen,Enas Abdulhay

doi:10.1155/2021/5359084

Abstract

The study aimed to explore the diagnostic value of computed tomography (CT) images based on cavity convolution U-Net algorithm for patients with severe pulmonary infection. A new lung CT image segmentation algorithm (U-Net+ deep convolution (DC)) was proposed based on U-Net network and compared with convolutional neural network (CNN) algorithm. Then, it was applied to CT image diagnosis of 100 patients with severe lung infection in The Second Affiliated Hospital of Fujian Medical University hospital and compared with traditional methods, and its sensitivity, specificity, and accuracy were compared. It was found that the single training time and loss of U-Net + DC algorithm were reduced by 59.4% and 9.8%, respectively, compared with CNN algorithm, while Dice increased by 3.6%. The lung contour segmented by the proposed model was smooth, which was the closest to the gold standard. Fungal infection, bacterial infection, viral infection, tuberculosis infection, and mixed infection accounted for 28%, 18%, 7%, 7%, and 40%, respectively. 36%, 38%, 26%, 17%, and 20% of the patients had ground-glass shadow, solid shadow, nodule or mass shadow, reticular or linear shadow, and hollow shadow in CT, respectively. The incidence of various CT characteristics in patients with fungal and bacterial infections was statistically significant (P < 0.05). The specificity (94.32%) and accuracy (97.22%) of CT image diagnosis based on U-Net + DC algorithm were significantly higher than traditional diagnostic method (75.74% and 74.23%), and the differences were statistically significant (P < 0.05). The network of the algorithm in this study demonstrated excellent image segmentation effect. The CT image based on the U-Net + DC algorithm can be used for the diagnosis of patients with severe pulmonary infection, with high diagnostic value.

Highlights

IntroductionThe incidence of postoperative pulmonary infection is very high [1]. If the diagnosis is not timely, it is easy to miss the best treatment time
At present, the incidence of postoperative pulmonary infection is very high [1]
It was noted that the single training time of the CNN algorithm was 690 s, and the single training time of the U-Net + deep convolution (DC) algorithm was 280 s. e single training time of the U-Net + DC algorithm was significantly less than the single training time of the CNN algorithm by 59.4%, and the difference was statistically significant (P < 0.05)

Summary

Introduction

The incidence of postoperative pulmonary infection is very high [1]. If the diagnosis is not timely, it is easy to miss the best treatment time. Studies have shown that if the diagnosis can be made in advance, the survival rate of patients with pulmonary infection can be increased by about 40% [2, 3]. Erefore, early and accurate diagnosis of pulmonary infection has become a key issue to improve the cure rate [4, 5]. In addition to pathogen testing, computed tomography (CT) imaging diagnosis has become a common and important method for medical diagnosis. As the main feature of judging pulmonary infection, lung CT signs are important from the clarity of lung images and the accurate expression of lung information. Traditional medical image segmentation extracts lung information according to the shallow features of the image and depends on the judgment of the doctor [6]

Methods

Results

Discussion

Conclusion