Abstract

For stage-I lung adenocarcinoma, the 5-years disease-free survival (DFS) rates of non-invasive adenocarcinoma (non-IA) is different with invasive adenocarcinoma (IA). This study aims to develop CT image based artificial intelligence (AI) schemes to classify between non-IA and IA nodules, and incorporate deep learning (DL) and radiomics features to improve the classification performance. We collect 373 surgical pathological confirmed ground-glass nodules (GGNs) from 323 patients in two centers. It involves 205 non-IA (including 107 adenocarcinoma in situ and 98 minimally invasive adenocarcinoma), and 168 IA. We first propose a recurrent residual convolutional neural network based on U-Net to segment the GGNs. Then, we build two schemes to classify between non-IA and IA namely, DL scheme and radiomics scheme, respectively. Third, to improve the classification performance, we fuse the prediction scores of two schemes by applying an information fusion method. Finally, we conduct an observer study to compare our scheme performance with two radiologists by testing on an independent dataset. Comparing with DL scheme and radiomics scheme (the area under a receiver operating characteristic curve (AUC): 0.83 ± 0.05, 0.87 ± 0.04), our new fusion scheme (AUC: 0.90 ± 0.03) significant improves the risk classification performance (p < 0.05). In a comparison with two radiologists, our new model yields higher accuracy of 80.3%. The kappa value for inter-radiologist agreement is 0.6. It demonstrates that applying AI method is an effective way to improve the invasiveness risk prediction performance of GGNs. In future, fusion of DL and radiomics features may have a potential to handle the classification task with limited dataset in medical imaging.

Highlights

  • As the most common histologic subtype of lung cancer, lung adenocarcinomas accounts for almost half of lung cancers

  • The diameters of 189 (50.7%) ground-glass nodules (GGNs) were smaller than 10 mm, the diameters of 148 (39.7%) GGNs were in a range of (10 mm, 20 mm), and the diameters of 36 (9.6%) GGNs were larger than 20 mm (P < 0.05)

  • We developed an artificial intelligence (AI) scheme to classify between non-invasive adenocarcinoma (IA) and IA GGNs in computed tomography (CT) images

Read more

Summary

Introduction

As the most common histologic subtype of lung cancer, lung adenocarcinomas accounts for almost half of lung cancers. The persistent presence of ground-glass nodules (GGN) in computed tomography (CT) image usually serves as an indicator of the presence of lung adenocarcinoma or its precursors [1]. According to the guideline of the 2011 International Association for the Predict Invasiveness of Lung Adenocarcinomas. Previous reported studies has depicted that the different subtypes of lung adenocarcinoma have different 3-years and 5-years disease-free survival (DFS) rates [3]. For stage-I lung adenocarcinoma, the 5-years DFS of AIS and MIA is 100%, but IA is only 38–86% [4, 5]. It is important to discriminate between IA and non-IA (including AIS and MIA) by using non-invasive CT image

Objectives
Methods
Results
Discussion
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call