Segmentation of Non-Small Cell Lung Carcinomas: Introducing DRU-Net and Multi-Lens Distortion
Authors:
- Soroush OskoueiSOSoroush Oskouei
- Marit VallaMVMarit Valla
- André PedersenAPAndré Pedersen
- Erik SmistadESErik Smistad
- Vibeke Grotnes DaleVDVibeke Grotnes Dale
- Maren HøibøMHMaren Høibø
- Sissel Gyrid Freim WahlSWSissel Gyrid Freim Wahl
- Mats Dehli HaugumMHMats Dehli Haugum
- Thomas LangøTLThomas Langø
- Maria Paula RamnefjellMRMaria Paula Ramnefjell
- Lars Andreas AkslenLALars Andreas Akslen
- Gabriel KissGKGabriel Kiss
Open Access
- 20 May 2025
- journal article
- research article
- Published by MDPI AG in Journal of Imaging
- Vol. 11 (5) , 166
Abstract
The increased workload in pathology laboratories today means automated tools such as artificial intelligence models can be useful, helping pathologists with their tasks. In this paper, we propose a segmentation model (DRU-Net) that can provide a delineation of human non-small cell lung carcinomas and an augmentation method that can improve classification results. The proposed model is a fused combination of truncated pre-trained DenseNet201 and ResNet101V2 as a patch-wise classifier, followed by a lightweight U-Net as a refinement model. Two datasets (Norwegian Lung Cancer Biobank and Haukeland University Lung Cancer cohort) were used to develop the model. The DRU-Net model achieved an average of 0.91 Dice similarity coefficient. The proposed spatial augmentation method (multi-lens distortion) improved the Dice similarity coefficient from 0.88 to 0.91. Our findings show that selecting image patches that specifically include regions of interest leads to better results for the patch-wise classifier compared to other sampling methods. A qualitative analysis by pathology experts showed that the DRU-Net model was generally successful in tumor detection. Results in the test set showed some areas of false-positive and false-negative segmentation in the periphery, particularly in tumors with inflammatory and reactive changes. In summary, the presented DRU-Net model demonstrated the best performance on the segmentation task, and the proposed augmentation technique proved to improve the results.Keywords
Funding Information
- The Liaison Committee for education, research, and innovation in Central Norway (2021/928, 2022/787, 223250)
- Norges Forskningsråd (2021/928, 2022/787, 223250)
This publication has 69 references indexed in Scilit:
- DEPICTER: Deep representation clustering for histology annotationComputers in Biology and Medicine, 2024
- Comparing CNN-based and transformer-based models for identifying lung cancer: which is more effective?Multimedia Tools and Applications, 2023
- DHUnet: Dual-branch hierarchical global–local fusion network for whole slide image segmentationBiomedical Signal Processing and Control, 2023
- MAMC-Net: an effective deep learning framework for whole-slide image tumor segmentationMultimedia Tools and Applications, 2023
- The state of the art for artificial intelligence in lung digital pathologyThe Journal of Pathology, 2022
- Deep contrastive learning based tissue clustering for annotation-free histopathology image analysisComputerized Medical Imaging and Graphics, 2022
- Deep transfer learning based model for colorectal cancer histopathology segmentation: A comparative study of deep pre-trained modelsInternational Journal of Medical Informatics, 2022
- Deep Multi-Magnification Networks for multi-class breast cancer image segmentationComputerized Medical Imaging and Graphics, 2021
- FAST: framework for heterogeneous medical image computing and visualizationInternational Journal of Computer Assisted Radiology and Surgery, 2015
- The 2015 WHO classification of lung tumorsDer Pathologe, 2014