Lung nodule malignancy prediction is an essential step in the early diagnosis of lung cancer. Besides the difficulties commonly discussed, the challenges of this task also come from the ambiguous labels provided by annotators, since deep learning models have in some cases been found to reproduce or amplify human biases. In this paper, we propose a multi-view 'divide-and-rule' (MV-DAR) model to learn from both reliable and ambiguous annotations for lung nodule malignancy prediction on chest CT scans. According to the consistency and reliability of their annotations, we divide nodules into three sets: a consistent and reliable set (CR-Set), an inconsistent set (IC-Set), and a low reliable set (LR-Set). The nodule in IC-Set is annotated by multiple radiologists inconsistently, and the nodule in LR-Set is annotated by only one radiologist. Although ambiguous, inconsistent labels tell which label(s) is consistently excluded by all annotators, and the unreliable labels of a cohort of nodules are largely correct from the statistical point of view. Hence, both IC-Set and LR-Set can be used to facilitate the training of MV-DAR. Our MV-DAR contains three DAR models to characterize a lung nodule from three orthographic views and is trained following a two-stage procedure. Each DAR consists of three networks with the same architecture, including a prediction network (Prd-Net), a counterfactual network (CF-Net), and a low reliable network (LR-Net), which are trained on CR-Set, IC-Set, and LR-Set respectively in the pretraining phase. In the fine-tuning phase, the image representation ability learned by CF-Net and LR-Net is transferred to Prd-Net by negative-attention module (NA-Module) and consistent-attention module (CA-Module), aiming to boost the prediction ability of Prd-Net. The MV-DAR model has been evaluated on the LIDC-IDRI dataset and LUNGx dataset. Our results indicate not only the effectiveness of the MV-DAR in learning from ambiguous labels but also its superiority over present noisy label-learning models in lung nodule malignancy prediction.
Download full-text PDF |
Source |
---|---|
http://dx.doi.org/10.1109/TMI.2022.3149344 | DOI Listing |
Clin Lung Cancer
January 2025
Thoracic Surgery Unit, IRCCS National Cancer Institute Regina Elena, Rome, Italy.
Introduction: To analyze the impact of Kirsten-Rat-Sarcoma Virus (KRAS) mutations on tumor-growth as estimated by tumor-doubling-time (TDT) among solid-dominant clinical-stage I lung adenocarcinoma. Moreover, to evaluate the prognostic role of KRAS mutations, TDT and their combination in completely-resected pathologic-stage I adenocarcinomas.
Methods: In this single-center retrospective analysis, completely resected clinical-stage I adenocarcinomas presenting as solid-dominant nodules (consolidation-to-tumor ratio > 0.
BMJ Open
January 2025
Centre for Cancer Screening, Prevention and Early Diagnosis, Wolfson Institute of Population Health, Barts and The London School of Medicine and Dentistry, Queen Mary University of London, London, UK.
Background: Worldwide, lung cancer (LC) is the second most frequent cancer and the leading cause of cancer related mortality. Low-dose CT (LDCT) screening reduced LC mortality by 20-24% in randomised trials of high-risk populations. A significant proportion of those screened have nodules detected that are found to be benign.
View Article and Find Full Text PDFJ Clin Med
January 2025
Translational Research Unit, Hospital Universitario Miguel Servet, IIS Aragón, 50009 Zaragoza, Spain.
Lung cancer is the primary cause of cancer-related deaths. Most patients are typically diagnosed at advanced stages. Low-dose computed tomography (LDCT) has been proven to reduce lung cancer mortality, but screening programs using LDCT are associated with a high number of false positives and unnecessary thoracotomies.
View Article and Find Full Text PDFLife (Basel)
December 2024
Department of Functional Science, "Victor Babes" University of Medicine and Pharmacy, 300041 Timisoara, Romania.
Background And Objectives: Lung cancer screening is critical for early detection and management, particularly through the use of computed tomography (CT). This study aims to compare the Lung Imaging Reporting and Data System (Lung-RADS) Version 2022 with the British Thoracic Society (BTS) guidelines in classifying solid pulmonary nodules detected at lung cancer screening CT examinations.
Materials And Methods: This retrospective study included 224 patients who underwent lung cancer screening CT between 2016 and 2022 and had a reported solid pulmonary nodule.
Biomolecules
December 2024
Department of Translational Medicine, University of Ferrara, Via Aldo Moro 8, 44124 Ferrara, Italy.
Prostate cancer (PCa) is a high-prevalence disease usually characterized by metastatic spread to the pelvic lymph nodes and bones and the development of visceral metastases only in the late stages of disease. Positron Emission Tomography (PET) plays a key role in the detection of PCa metastases. Several PET radiotracers are used in PCa patients according to the stage and pathological features of the disease, in particular Ga/F-prostate-specific membrane antigen (PSMA) ligands.
View Article and Find Full Text PDFEnter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!