A novel multi-scale CNNs for false positive reduction in pulmonary nodule detection

作者：

Highlights：

• To shorten the training time, three orthogonal slices are cropped from 3D CT cubes.

• The three orthogonal slices are stacked to preserve the nodules spatial information.

• Spatial and frequency domain information are used to extract significant features.

• A new multi-scale framework is designed to better encode contextual information.

摘要

•To shorten the training time, three orthogonal slices are cropped from 3D CT cubes.•The three orthogonal slices are stacked to preserve the nodules spatial information.•Spatial and frequency domain information are used to extract significant features.•A new multi-scale framework is designed to better encode contextual information.

论文关键词：False positive reduction,Multi-scale CNNs,Lung nodule

论文评审过程：Received 11 September 2021, Revised 28 April 2022, Accepted 27 May 2022, Available online 6 June 2022, Version of Record 30 June 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.117652