Classification of Lung Nodule Malignancy Risk on Computed Tomography Images Using Convolutional Neural Network: A Comparison Between 2D and 3D Strategies

Yan, Xingjian; Pang, Jianing; Hu, Qi; Zhu, Yixin; Bai, Chunxue; Geng, Xin; Liu, Mina; Terzopoulos, Demetri; Ding, Xiaowei

doi:10.1007/978-3-319-54526-4_7

Cited by 59 publications

(55 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Deep transfer learning and multiinstance learning is used for patient-level lung CT diagnosis [25,36]. A comparative study on 2D and 3D ConvNets is conducted and 3D ConvNet is shown to be better than 2D ConvNet for 3D CT data [33]. Furthermore, a multitask learning and transfer learning framework is proposed for nodule diagnosis [14].…”

Section: Related Workmentioning

confidence: 99%

“…Accuracy (%) Year Multi-scale CNN [26] 86.84 2015 Slice-level 2D CNN [33] 86.70 2016 Nodule-level 2D CNN [33] 87.30 2016 Vanilla 3D CNN [33] 87.40 2016 Multi-crop CNN [27] 87 [26], Vanilla 3D CNN [33] and Multi-crop CNN [27], because of the strong power of 3D structure and deep dual path network. GBM with nodule size and raw nodule pixels with crop size as 16 × 16 × 16 achieves comparable performance as multiscale CNN [26] because of the superior classification per-…”

Section: Modelsmentioning

confidence: 99%

See 1 more Smart Citation

DeepLung: Deep 3D Dual Path Nets for Automated Pulmonary Nodule Detection and Classification

Zhu

Liu

Wang

et al. 2018

2018 IEEE Winter Conference on Applications of Computer Vision (WACV)

331

220

View full text Add to dashboard Cite

In this work, we present a fully automated lung computed tomography (CT) cancer diagnosis system, DeepLung. DeepLung consists of two components, nodule detection (identifying the locations of candidate nodules) and classification (classifying candidate nodules into benign or malignant). Considering the 3D nature of lung CT data and the compactness of dual path networks (DPN), two deep 3D DPN are designed for nodule detection and classification respectively. Specifically, a 3D Faster Regions with Convolutional Neural Net (R-CNN) is designed for nodule detection with 3D dual path blocks and a U-net-like encoder-decoder structure to effectively learn nodule features. For nodule classification, gradient boosting machine (GBM) with 3D dual path network features is proposed. The nodule classification subnetwork was validated on a public dataset from LIDC-IDRI, on which it achieved better performance than state-of-the-art approaches and surpassed the performance of experienced doctors based on image modality. Within the DeepLung system, candidate nodules are detected first by the nodule detection subnetwork, and nodule diagnosis is conducted by the classification subnetwork. Extensive experimental results demonstrate that DeepLung has performance comparable to experienced doctors both for the nodule-level and patient-level diagnosis on the LIDC-IDRI dataset. 1

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Modelsmentioning

confidence: 99%

DeepLung: Deep 3D Dual Path Nets for Automated Pulmonary Nodule Detection and Classification

Zhu

Liu

Wang

et al. 2018

2018 IEEE Winter Conference on Applications of Computer Vision (WACV)

331

220

View full text Add to dashboard Cite

show abstract

“…Yan et al [13] proponen utilizar una red neuronal convolucional para clasificar nódulos pulmonares (en benignos y malignos), así mismo, implementan tres redes neuronales: (i) 2D a nivel de división (ii) 2D a nivel de nódulos (iii) 3D a nivel de nódulos. Para comparar las redes neuronales utilizan el dataset LIDC-IDRI, obteniendo mejores resultados en la red 3D a nivel de nódulos.…”

Section: Revisión De La Literaturaunclassified

“…Mediante el aprendizaje automático se pueden detectar y predecir patrones [4], ha mostrado ser efectivo en aplicaciones delámbito médico, tales como: detección temprana de enfermedades, planificación quirúrgica, análisis de patologías, oncología y radiología. El aprendizaje automático es una de las disciplinas de la inteligencia artificial, que emplea estadística, probabilidad y técnicas de optimización, para comprender reglas complicadas y características de alta dimensión; en el tema de cáncer de pulmón se ha aplicado para detectarlo [5], clasificar la gravedad del cáncer [6], predecir la supervivencia de pacientes [7], clasificar el cáncer de pulmón [8,9] y clasificar nódulos pulmonares [10][11][12][13][14][15][16][17][18][19].…”

Section: Introductionunclassified

Estado del arte de clasificación de nódulos pulmonares utilizando Aprendizaje Automático

Mejía¹,

Biswal²,

Sánchez-DelaCruz³

2019

RCS

View full text Add to dashboard Cite

show abstract

“…Dual path connection integrates the advantages of the two advanced frameworks, residual learning for feature reuse and dense connection for keeping exploiting new features, into a unified structure, which obtains success on the Ima-geNet dataset. For CT data, advanced method should be effective to extract 3D volume feature (Yan et al 2016). We design a 3D deep dual path network for the 3D CT lung nodule classification in Fig.…”

Section: Gradient Boosting Machine With 3d Dual Path Net Feature For mentioning

confidence: 99%

DeepLung: 3D Deep Convolutional Nets for Automated Pulmonary Nodule Detection and Classification

Zhu

Liu

Wang

et al. 2017

Preprint

View full text Add to dashboard Cite

In this work, we present a fully automated lung CT cancer diagnosis system, DeepLung. DeepLung contains two parts, nodule detection and classification. Considering the 3D nature of lung CT data, two 3D networks are designed for the nodule detection and classification respectively. Specifically, a 3D Faster R-CNN is designed for nodule detection with a U-net-like encoder-decoder structure to effectively learn nodule features. For nodule classification, gradient boosting machine (GBM) with 3D dual path network (DPN) features is proposed. The nodule classification subnetwork is validated on a public dataset from LIDC-IDRI, on which it achieves better performance than state-of-the-art approaches, and surpasses the average performance of four experienced doctors. For the DeepLung system, candidate nodules are detected first by the nodule detection subnetwork, and nodule diagnosis is conducted by the classification subnetwork. Extensive experimental results demonstrate the DeepLung is comparable to the experienced doctors both for the nodule-level and patient-level diagnosis on the LIDC-IDRI dataset.

show abstract

Classification of Lung Nodule Malignancy Risk on Computed Tomography Images Using Convolutional Neural Network: A Comparison Between 2D and 3D Strategies

Cited by 59 publications

References 16 publications

DeepLung: Deep 3D Dual Path Nets for Automated Pulmonary Nodule Detection and Classification

DeepLung: Deep 3D Dual Path Nets for Automated Pulmonary Nodule Detection and Classification

Estado del arte de clasificación de nódulos pulmonares utilizando Aprendizaje Automático

DeepLung: 3D Deep Convolutional Nets for Automated Pulmonary Nodule Detection and Classification

Contact Info

Product

Resources

About