LN-DETR: cross-scale feature fusion and re-weighting for lung nodule detection

Abstract With advancements in technology, lung nodule detection has significantly improved in both speed and accuracy. However, challenges remain in deploying these methods in complex real-world scenarios. This paper introduces an enhanced lung nodule detection algorithm base on RT-DETR, called LN-D...

Full description

Saved in:
Bibliographic Details
Main Authors: Dibin Zhou, Honggang Xu, Wenhao Liu, Fuchang Liu
Format: Article
Language:English
Published: Nature Portfolio 2025-05-01
Series:Scientific Reports
Subjects:
Online Access:https://doi.org/10.1038/s41598-025-00309-7
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850206563023716352
author Dibin Zhou
Honggang Xu
Wenhao Liu
Fuchang Liu
author_facet Dibin Zhou
Honggang Xu
Wenhao Liu
Fuchang Liu
author_sort Dibin Zhou
collection DOAJ
description Abstract With advancements in technology, lung nodule detection has significantly improved in both speed and accuracy. However, challenges remain in deploying these methods in complex real-world scenarios. This paper introduces an enhanced lung nodule detection algorithm base on RT-DETR, called LN-DETR. First, we designed a Deep and Shallow Detail Fusion layer that effectively fuses cross-scale features from both shallow and deep layers. Second, we optimized the computational load of the backbone network, effectively reducing the overall scale of the model. Finally, an efficient downsampling is designed to enhance the detection of lung nodules by re-weighting contextual information. Experiments conducted on the public LUNA16 dataset demonstrate that the proposed method, with a reduced number of parameters and computational overhead, achieves 83.7% mAP@0.5 and 36.3% mAP@0.5:0.95, outperforming RT-DETR in both model size and accuracy. These results highlight the superior detection accuracy of the proposed network while maintaining computational efficiency.
format Article
id doaj-art-167a9be98faa4861af9957409e1608aa
institution OA Journals
issn 2045-2322
language English
publishDate 2025-05-01
publisher Nature Portfolio
record_format Article
series Scientific Reports
spelling doaj-art-167a9be98faa4861af9957409e1608aa2025-08-20T02:10:46ZengNature PortfolioScientific Reports2045-23222025-05-0115111810.1038/s41598-025-00309-7LN-DETR: cross-scale feature fusion and re-weighting for lung nodule detectionDibin Zhou0Honggang Xu1Wenhao Liu2Fuchang Liu3School of Information Science and Technology, Hangzhou Normal UniversitySchool of Information Science and Technology, Hangzhou Normal UniversitySchool of Information Science and Technology, Hangzhou Normal UniversitySchool of Information Science and Technology, Hangzhou Normal UniversityAbstract With advancements in technology, lung nodule detection has significantly improved in both speed and accuracy. However, challenges remain in deploying these methods in complex real-world scenarios. This paper introduces an enhanced lung nodule detection algorithm base on RT-DETR, called LN-DETR. First, we designed a Deep and Shallow Detail Fusion layer that effectively fuses cross-scale features from both shallow and deep layers. Second, we optimized the computational load of the backbone network, effectively reducing the overall scale of the model. Finally, an efficient downsampling is designed to enhance the detection of lung nodules by re-weighting contextual information. Experiments conducted on the public LUNA16 dataset demonstrate that the proposed method, with a reduced number of parameters and computational overhead, achieves 83.7% mAP@0.5 and 36.3% mAP@0.5:0.95, outperforming RT-DETR in both model size and accuracy. These results highlight the superior detection accuracy of the proposed network while maintaining computational efficiency.https://doi.org/10.1038/s41598-025-00309-7Deep learningObject detectionLung noduleMedical image processing
spellingShingle Dibin Zhou
Honggang Xu
Wenhao Liu
Fuchang Liu
LN-DETR: cross-scale feature fusion and re-weighting for lung nodule detection
Scientific Reports
Deep learning
Object detection
Lung nodule
Medical image processing
title LN-DETR: cross-scale feature fusion and re-weighting for lung nodule detection
title_full LN-DETR: cross-scale feature fusion and re-weighting for lung nodule detection
title_fullStr LN-DETR: cross-scale feature fusion and re-weighting for lung nodule detection
title_full_unstemmed LN-DETR: cross-scale feature fusion and re-weighting for lung nodule detection
title_short LN-DETR: cross-scale feature fusion and re-weighting for lung nodule detection
title_sort ln detr cross scale feature fusion and re weighting for lung nodule detection
topic Deep learning
Object detection
Lung nodule
Medical image processing
url https://doi.org/10.1038/s41598-025-00309-7
work_keys_str_mv AT dibinzhou lndetrcrossscalefeaturefusionandreweightingforlungnoduledetection
AT honggangxu lndetrcrossscalefeaturefusionandreweightingforlungnoduledetection
AT wenhaoliu lndetrcrossscalefeaturefusionandreweightingforlungnoduledetection
AT fuchangliu lndetrcrossscalefeaturefusionandreweightingforlungnoduledetection