Multi-level 3D Densenets for False-positive Reduction in Lung Nodule Detection Based on Chest Computed Tomography

2020 ◽  
Vol 16 (8) ◽  
pp. 1004-1021
Author(s):  
Xiaoqi Lu ◽  
Yu Gu ◽  
Lidong Yang ◽  
Baohua Zhang ◽  
Ying Zhao ◽  
...  
Keyword(s):  
False Positive ◽  
Contextual Information ◽  
Lung Nodule ◽  
Positive Sample ◽  
Lung Nodules ◽  
Lung Nodule Detection ◽  
Nodule Detection ◽  
Level Information ◽  
Multi Level ◽  
Cade System

Objective: False-positive nodule reduction is a crucial part of a computer-aided detection (CADe) system, which assists radiologists in accurate lung nodule detection. In this research, a novel scheme using multi-level 3D DenseNet framework is proposed to implement false-positive nodule reduction task. Methods: Multi-level 3D DenseNet models were extended to differentiate lung nodules from falsepositive nodules. First, different models were fed with 3D cubes with different sizes for encoding multi-level contextual information to meet the challenges of the large variations of lung nodules. In addition, image rotation and flipping were utilized to upsample positive samples which consisted of a positive sample set. Furthermore, the 3D DenseNets were designed to keep low-level information of nodules, as densely connected structures in DenseNet can reuse features of lung nodules and then boost feature propagation. Finally, the optimal weighted linear combination of all model scores obtained the best classification result in this research. Results: The proposed method was evaluated with LUNA16 dataset which contained 888 thin-slice CT scans. The performance was validated via 10-fold cross-validation. Both the Free-response Receiver Operating Characteristic (FROC) curve and the Competition Performance Metric (CPM) score show that the proposed scheme can achieve a satisfactory detection performance in the falsepositive reduction track of the LUNA16 challenge. Conclusion: The result shows that the proposed scheme can be significant for false-positive nodule reduction task.

A multi-level file structure for information processing

1960 ◽  
Author(s):  
L. Miller ◽  
J. Minker ◽  
W. G. Reed ◽  
W. E. Shindle
Keyword(s):  
Information Processing ◽  
File Structure ◽  
Multi Level

scholarly journals Neo-mechanistic explanatory integration for cognitive science: the problem of reduction remains

Sofia ◽  
2019 ◽  
Vol 8 (1) ◽  
pp. 124-145 ◽  
Author(s):  
Diego Azevedo Leite
Keyword(s):  
Information Processing ◽  
Cognitive Science ◽  
Cognitive Models ◽  
Human Cognition ◽  
Cognitive Mechanisms ◽  
Cognitive Sciences ◽  
Significant Challenge ◽  
Neural Processes ◽  
Cognitive Theories ◽  
Multi Level

One of the central aims of the neo-mechanistic framework for the neural and cognitive sciences is to construct a pluralistic integration of scientific explanations, allowing for a weak explanatory autonomy of higher-level sciences, such as cognitive science. This integration involves understanding human cognition as information processing occurring in multi-level human neuro-cognitive mechanisms, explained by multi-level neuro-cognitive models. Strong explanatory neuro-cognitive reduction, however, poses a significant challenge to this pluralist ambition and the weak autonomy of cognitive science derived therefrom. Based on research in current molecular and cellular neuroscience, the framework holds that the best strategy for integrating human neuro-cognitive theories is through direct reductive explanations based on molecular and cellular neural processes. It is my aim to investigate whether the neo-mechanistic framework can meet the challenge. I argue that leading neo-mechanists offer some significant replies; however, they are not able yet to completely remove strong explanatory reductionism from their own framework.

Dual Attention Multiscale Network for Vessel Segmentation in Fundus Photography

2021 ◽  
Author(s):  
Pengshuai Yin ◽  
Yupeng Fang ◽  
Qingyao Wu ◽  
QiLin Wan
Keyword(s):  
Disease Diagnosis ◽  
Fundus Photography ◽  
Essential Step ◽  
Input Size ◽  
Real World Applications ◽  
Vessel Structure ◽  
Level Information ◽  
Multi Level ◽  
Sensitivity Specificity ◽  
To Receive

Abstract Background: Automatic vessel structure segmentation is an essential step towards an automatic disease diagnosis system. The task is challenging due to the variance shapes and sizes of vessels across populations.Methods: A multiscale network with dual attention is proposed to segment vessels in different sizes. The network injects spatial attention module and channel attention module on feature map which size is 1 8 of the input size. The network also uses multiscale input to receive multi-level information, and the network uses the multiscale output to gain more supervision. Results: The proposed method is tested on two publicly available datasets: DRIVE and CHASEDB1. The accuracy, AUC, sensitivity, specificity on DRIVE dataset is 0.9615, 0.9866, 0.7693, and 0.9851, respectively. On the CHASEDB1 dataset, the metrics are 0.9797, 0.9895, 0.8432, and 0.9863 respectively. The ablative study further shows effectiveness for each part of the network. Conclusions: Multiscale and dual attention mechanism both improves the performance. The proposed architecture is simple and effective. The inference time is 12ms on a GPU and has potential for real-world applications. The code will be made publicly available.

Quantum Wavelet Transforms

2021 ◽  
pp. 193-220
Keyword(s):  
Information Processing ◽  
Quantum Information ◽  
Wavelet Transform ◽  
Tensor Product ◽  
Wavelet Transforms ◽  
Quantum Information Processing ◽  
Complexity Analysis ◽  
Simulation Experiments ◽  
Multi Level

The classical wavelet transform has been widely applied in the information processing field. It implies that quantum wavelet transform (QWT) may play an important role in quantum information processing. This chapter firstly describes the iteration equations of the general QWT using generalized tensor product. Then, Haar QWT (HQWT), Daubechies D4 QWT (DQWT), and their inverse transforms are proposed respectively. Meanwhile, the circuits of the two kinds of multi-level HQWT are designed. What's more, the multi-level DQWT based on the periodization extension is implemented. The complexity analysis shows that the proposed multi-level QWTs on 2n elements can be implemented by O(n3) basic operations. Simulation experiments demonstrate that the proposed QWTs are correct and effective.

Application of multi-level information fusion for wear particle recognition of ferrographic images

2018 ◽  
Vol 26 (6) ◽  
pp. 1551-1560
Author(s):  
徐 斌 XU Bin ◽  
温广瑞 WEN Guang-rui ◽  
苏 宇 SU Yu ◽  
张志芬 ZHANG Zhi-fen ◽  
陈 峰 CHEN Feng ◽  
...  
Keyword(s):  
Information Fusion ◽  
Wear Particle ◽  
Particle Recognition ◽  
Level Information ◽  
Multi Level

Stimulating learning by empowering leadership

2016 ◽  
Vol 37 (8) ◽  
pp. 1168-1186 ◽  
Author(s):  
Mingze Li ◽  
Pengcheng Zhang
Keyword(s):  
Information Processing ◽  
Team Learning ◽  
Hierarchical Linear Model ◽  
Hierarchical Regression ◽  
Empowering Leadership ◽  
Content Type ◽  
Team Members ◽  
Section Data ◽  
Multi Level ◽  
Different Levels

Purpose The purpose of this paper is to answer the theoretical and practical calls for an examination of the multi-level effects of empowering leadership on creativity. In addition, it attempts to link empowering leadership to creativity from the perspective of information processing, which is different from traditional mechanisms of psychology. Design/methodology/approach Based on the perspective of information processing, the authors tested how and why different levels of empowering leadership may relate to team and individual creativity. Multi-source data were collected from 62 team leaders and 295 team members. Statistical methods, such as the hierarchical linear model, hierarchical regression analysis, and bootstrapping tests, were used to analyze the data. Findings The results show that team and individual learning mediate the effects of empowering leadership on creativity at the team and individual levels. Interestingly, the authors also found that team learning negatively moderates the indirect and positive effect of individual empowering leadership on individual creativity. Research limitations/implications The main limitation of this study is that the authors used cross-section data instead of longitudinal data to analyze the causal relationship. As such, the results may not truly reveal the causality. Practical implications The findings indicate that empowering leadership is important for stimulating both individual and team learning; thus, it benefits different levels of creativity. In addition, the results also suggest that there are interplay between different level mechanisms, and empowering team leader should trade-off individual and team learning effects in order to promote both team and individual creativity effectively. Originality/value This study contributes to the existing literature by providing a multi-level and cross-level analysis of empowering leadership and creativity. It clarifies how empowering leadership stimulates individual and team creativity at different levels simultaneously.

Sign in / Sign up

Export Citation Format

Share Document