Convolution kernel and iterative reconstruction affect the diagnostic performance of radiomics and deep learning in lung adenocarcinoma pathological subtypes

Autor: Yingli Sun, Liang Jin, Wufei Chen, Wei Zhao, Cheng Li, Jianying Li, Wei Zhang, Peijun Wang, Jiancheng Yang, Yanqing Hua, Yuxiang Ye, Pan Gao, Ming Li
Rok vydání: 2019
Předmět:
0301 basic medicine
Pulmonary and Respiratory Medicine
iterative reconstruction
Medical Informatics Computing
Adenocarcinoma of Lung
Iterative reconstruction
Logistic regression
lcsh:RC254-282
Convolution
03 medical and health sciences
0302 clinical medicine
Radiomics
Image Processing
Computer-Assisted
medicine
Humans
Neoplasm Staging
lung adenocarcinomas
Receiver operating characteristic
business.industry
Deep learning
deep learning
Pattern recognition
Original Articles
General Medicine
lcsh:Neoplasms. Tumors. Oncology. Including cancer and carcinogens
medicine.disease
030104 developmental biology
ROC Curve
Oncology
radiomics
Sample size determination
Area Under Curve
030220 oncology & carcinogenesis
Adenocarcinoma
Original Article
Artificial intelligence
Neoplasm Grading
Tomography
X-Ray Computed
Convolution kernel
business
Zdroj: Thoracic Cancer
Thoracic Cancer, Vol 10, Iss 10, Pp 1893-1903 (2019)
ISSN: 1759-7714
1759-7706
DOI: 10.1111/1759-7714.13161
Popis: Background The aim of this study was to investigate the influence of convolution kernel and iterative reconstruction on the diagnostic performance of radiomics and deep learning (DL) in lung adenocarcinomas. Methods A total of 183 patients with 215 lung adenocarcinomas were included in this study. All CT imaging data was reconstructed with three reconstruction algorithms (ASiR at 0%, 30%, 60% strength), each with two convolution kernels (bone and standard). A total of 171 nodules were selected as the training‐validation set, whereas 44 nodules were selected as the testing set. Logistic regression and a DL framework‐DenseNets were selected to tackle the task. Three logical experiments were implemented to fully explore the influence of the studied parameters on the diagnostic performance. The receiver operating characteristic curve (ROC) was used to evaluate the performance of constructed models. Results In Experiments A and B, no statistically significant results were found in the radiomic method, whereas two and six pairs were statistically significant (P
Databáze: OpenAIRE
Externí odkaz:
Nepřihlášeným uživatelům se plný text nezobrazuje