This website requires JavaScript.

MEDS-Net: Self-Distilled Multi-Encoders Network with Bi-Direction Maximum Intensity projections for Lung Nodule Detection

Muhammad UsmanAzka RehmanAbdullah Shahid ...+5 Yeong Gil Shin
Oct 2022
摘要
In this study, we propose a lung nodule detection scheme which fullyincorporates the clinic workflow of radiologists. Particularly, we exploitBi-Directional Maximum intensity projection (MIP) images of various thicknesses(i.e., 3, 5 and 10mm) along with a 3D patch of CT scan, consisting of 10adjacent slices to feed into self-distillation-based Multi-Encoders Network(MEDS-Net). The proposed architecture first condenses 3D patch input to threechannels by using a dense block which consists of dense units which effectivelyexamine the nodule presence from 2D axial slices. This condensed information,along with the forward and backward MIP images, is fed to three differentencoders to learn the most meaningful representation, which is forwarded intothe decoded block at various levels. At the decoder block, we employ aself-distillation mechanism by connecting the distillation block, whichcontains five lung nodule detectors. It helps to expedite the convergence andimproves the learning ability of the proposed architecture. Finally, theproposed scheme reduces the false positives by complementing the main detectorwith auxiliary detectors. The proposed scheme has been rigorously evaluated on888 scans of LUNA16 dataset and obtained a CPM score of 93.6\%. The resultsdemonstrate that incorporating of bi-direction MIP images enables MEDS-Net toeffectively distinguish nodules from surroundings which help to achieve thesensitivity of 91.5% and 92.8% with false positives rate of 0.25 and 0.5 perscan, respectively.
展开全部
图表提取

暂无人提供速读十问回答

论文十问由沈向洋博士提出,鼓励大家带着这十个问题去阅读论文,用有用的信息构建认知模型。写出自己的十问回答,还有机会在当前页面展示哦。

Q1论文试图解决什么问题?
Q2这是否是一个新的问题?
Q3这篇文章要验证一个什么科学假设?
0
被引用
笔记
问答