HR-MPF: high-resolution representation network with multi-scale progressive fusion for pulmonary nodule segmentation and classification

Abstract Accurate segmentation and classification of pulmonary nodules are of great significance to early detection and diagnosis of lung diseases, which can reduce the risk of developing lung cancer and improve patient survival rate. In this paper, we propose an effective network for pulmonary nodu...

Descripción completa

Guardado en:
Detalles Bibliográficos
Autores principales: Ling Zhu, Hongqing Zhu, Suyi Yang, Pengyu Wang, Yang Yu
Formato: article
Lenguaje:EN
Publicado: SpringerOpen 2021
Materias:
Acceso en línea:https://doaj.org/article/e046543880dd48388f1df46816893c73
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
id oai:doaj.org-article:e046543880dd48388f1df46816893c73
record_format dspace
spelling oai:doaj.org-article:e046543880dd48388f1df46816893c732021-11-14T12:08:44ZHR-MPF: high-resolution representation network with multi-scale progressive fusion for pulmonary nodule segmentation and classification10.1186/s13640-021-00574-21687-5281https://doaj.org/article/e046543880dd48388f1df46816893c732021-11-01T00:00:00Zhttps://doi.org/10.1186/s13640-021-00574-2https://doaj.org/toc/1687-5281Abstract Accurate segmentation and classification of pulmonary nodules are of great significance to early detection and diagnosis of lung diseases, which can reduce the risk of developing lung cancer and improve patient survival rate. In this paper, we propose an effective network for pulmonary nodule segmentation and classification at one time based on adversarial training scheme. The segmentation network consists of a High-Resolution network with Multi-scale Progressive Fusion (HR-MPF) and a proposed Progressive Decoding Module (PDM) recovering final pixel-wise prediction results. Specifically, the proposed HR-MPF firstly incorporates boosted module to High-Resolution Network (HRNet) in a progressive feature fusion manner. In this case, feature communication is augmented among all levels in this high-resolution network. Then, downstream classification module would identify benign and malignant pulmonary nodules based on feature map from PDM. In the adversarial training scheme, a discriminator is set to optimize HR-MPF and PDM through back propagation. Meanwhile, a reasonably designed multi-task loss function optimizes performance of segmentation and classification overall. To improve the accuracy of boundary prediction crucial to nodule segmentation, a boundary consistency constraint is designed and incorporated in the segmentation loss function. Experiments on publicly available LUNA16 dataset show that the framework outperforms relevant advanced methods in quantitative evaluation and visual perception.Ling ZhuHongqing ZhuSuyi YangPengyu WangYang YuSpringerOpenarticlePulmonary noduleSegmentation and classificationHigh-resolution networkMulti-scale progressive fusionGenerative adversarial networkElectronicsTK7800-8360ENEURASIP Journal on Image and Video Processing, Vol 2021, Iss 1, Pp 1-26 (2021)
institution DOAJ
collection DOAJ
language EN
topic Pulmonary nodule
Segmentation and classification
High-resolution network
Multi-scale progressive fusion
Generative adversarial network
Electronics
TK7800-8360
spellingShingle Pulmonary nodule
Segmentation and classification
High-resolution network
Multi-scale progressive fusion
Generative adversarial network
Electronics
TK7800-8360
Ling Zhu
Hongqing Zhu
Suyi Yang
Pengyu Wang
Yang Yu
HR-MPF: high-resolution representation network with multi-scale progressive fusion for pulmonary nodule segmentation and classification
description Abstract Accurate segmentation and classification of pulmonary nodules are of great significance to early detection and diagnosis of lung diseases, which can reduce the risk of developing lung cancer and improve patient survival rate. In this paper, we propose an effective network for pulmonary nodule segmentation and classification at one time based on adversarial training scheme. The segmentation network consists of a High-Resolution network with Multi-scale Progressive Fusion (HR-MPF) and a proposed Progressive Decoding Module (PDM) recovering final pixel-wise prediction results. Specifically, the proposed HR-MPF firstly incorporates boosted module to High-Resolution Network (HRNet) in a progressive feature fusion manner. In this case, feature communication is augmented among all levels in this high-resolution network. Then, downstream classification module would identify benign and malignant pulmonary nodules based on feature map from PDM. In the adversarial training scheme, a discriminator is set to optimize HR-MPF and PDM through back propagation. Meanwhile, a reasonably designed multi-task loss function optimizes performance of segmentation and classification overall. To improve the accuracy of boundary prediction crucial to nodule segmentation, a boundary consistency constraint is designed and incorporated in the segmentation loss function. Experiments on publicly available LUNA16 dataset show that the framework outperforms relevant advanced methods in quantitative evaluation and visual perception.
format article
author Ling Zhu
Hongqing Zhu
Suyi Yang
Pengyu Wang
Yang Yu
author_facet Ling Zhu
Hongqing Zhu
Suyi Yang
Pengyu Wang
Yang Yu
author_sort Ling Zhu
title HR-MPF: high-resolution representation network with multi-scale progressive fusion for pulmonary nodule segmentation and classification
title_short HR-MPF: high-resolution representation network with multi-scale progressive fusion for pulmonary nodule segmentation and classification
title_full HR-MPF: high-resolution representation network with multi-scale progressive fusion for pulmonary nodule segmentation and classification
title_fullStr HR-MPF: high-resolution representation network with multi-scale progressive fusion for pulmonary nodule segmentation and classification
title_full_unstemmed HR-MPF: high-resolution representation network with multi-scale progressive fusion for pulmonary nodule segmentation and classification
title_sort hr-mpf: high-resolution representation network with multi-scale progressive fusion for pulmonary nodule segmentation and classification
publisher SpringerOpen
publishDate 2021
url https://doaj.org/article/e046543880dd48388f1df46816893c73
work_keys_str_mv AT lingzhu hrmpfhighresolutionrepresentationnetworkwithmultiscaleprogressivefusionforpulmonarynodulesegmentationandclassification
AT hongqingzhu hrmpfhighresolutionrepresentationnetworkwithmultiscaleprogressivefusionforpulmonarynodulesegmentationandclassification
AT suyiyang hrmpfhighresolutionrepresentationnetworkwithmultiscaleprogressivefusionforpulmonarynodulesegmentationandclassification
AT pengyuwang hrmpfhighresolutionrepresentationnetworkwithmultiscaleprogressivefusionforpulmonarynodulesegmentationandclassification
AT yangyu hrmpfhighresolutionrepresentationnetworkwithmultiscaleprogressivefusionforpulmonarynodulesegmentationandclassification
_version_ 1718429438074421248