Applying a CT texture analysis model trained with deep‐learning reconstruction images to iterative reconstruction images in pulmonary nodule diagnosis

Abstract

Objective To investigate the feasibility and accuracy of applying a computed tomography (CT) texture analysis model trained with deep‐learning reconstruction images to iterative reconstruction images for classifying pulmonary nodules. Materials and methods CT images of 102 patients, with a total of 118 pulmonary nodules (52 benign, 66 malignant) were retrospectively reconstructed with a deep‐learning reconstruction (artificial intelligence iterative reconstruction [AIIR]) and a hybrid iterative reconstruction (HIR) technique. The AIIR data were divided into a training (n = 96) and a validation set (n = 22), and the HIR data were set as the test set (n = 118). Extracted texture features were compared using the Mann‐Whitney U test and t‐test. The diagnostic performance of the classification model was analyzed with the receiver operating characteristic curve (ROC), the area under ROC (AUC), sensitivity, specificity, and accuracy. Results Among the obtained 68 texture features, 51 (75.0%) were not influenced by the change of reconstruction algorithm (p > 0.05). Forty‐four features were significantly different between benign and malignant nodules (p < 0.05) for the AIIR dataset, which were selected to build the classification model. The accuracy and AUC of the classification model were 92.3% and 0.91 (95% confidence interval [CI], 0.74−0.90) with the validation set, which were 80.0% and 0.80 (95% CI, 0.68−0.86) with the test set. Conclusion With the CT texture analysis model trained with deep‐learning reconstruction (AIIR) images showing favorable diagnostic accuracy in discriminating benign and malignant pulmonary nodules, it also has certain applicability to the iterative reconstruction (HIR) images.

Artificial Intelligence Iterative Reconstruction in Computed Tomography Angiography: An Evaluation on Pulmonary Arteries and Aorta With Routine Dose Settings

“…Pulmonary and aortic CTA scans were reconstructed using FBP, a routinely available HIR algorithm (Karl 3D, United Imaging Healthcare), and AIIR. Specifically, AIIR [10][11][12] is a CT reconstruction algorithm with the deep learning technique integrated into the model-based iterative reconstruction. 14 In developing this algorithm, the traditional regularization term in model-based iterative reconstruction for reducing noises and artifacts is replaced by a dedicated denoising convolutional neural network, which is trained with a large number of image pairs, ie, standard dose images (noise-free images) and the corresponding simulated low-dose images (noisy images).…”

“…The purpose of this study is to investigate whether a newly introduced deep learning-based iterative reconstruction algorithm, ie, the artificial intelligence iterative reconstruction (AIIR; United Imaging Healthcare, Shanghai, China), [10][11][12] has a value in CTA image quality, especially on the visualization of vascular structures and associated lesions with routine dose settings. If it is so, how exactly is the improvement compared with the filtered back projection (FBP) 13 and hybrid iterative reconstruction (HIR).…”