Endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) is a cornerstone in minimally invasive thoracic lymph node sampling. In lung cancer staging, precise assessment of lymph node position is crucial for clinical decision-making. This study aimed to demonstrate a new deep learning method to classify thoracic lymph nodes based on their anatomical location using EBUS images. Bronchoscopists labeled lymph node stations in real-time according to the Mountain Dressler nomenclature. EBUS images were then used to train and test a deep neural network (DNN) model, with intraoperative labels as ground truth. In total, 28,134 EBUS images were acquired from 56 patients. The model achieved an overall classification accuracy of 59.5 ± 5.2%. The highest precision, sensitivity, and F1 score were observed in station 4L, 77.6 ± 13.1%, 77.6 ± 15.4%, and 77.6 ± 15.4%, respectively. The lowest precision, sensitivity, and F1 score were observed in station 10L. The average processing and prediction time for a sequence of ten images was 0.65 ± 0.04 s, demonstrating the feasibility of real-time applications. In conclusion, the new DNN-based model could be used to classify lymph node stations from EBUS images. The method performance was promising with a potential for clinical use.
Accurate localization of thoracic lymph nodes during endobronchial ultrasound (EBUS) is crucial for lung cancer staging, treatment planning, and prognostication. Artificial intelligence (AI) has the potential to support this process. Deep learning (DL) models often lack transparency but can benefit from explainable AI (XAI) tools like Gradient-weighted Class Activation Mapping (Grad-CAM). However, no prior study has quantitatively assessed whether model attention in EBUS imaging corresponds to relevant anatomy. This study developed a convolutional neural network (CNN) to classify thoracic lymph node stations and evaluated the anatomical relevance of Grad-CAM activations using a structured annotation framework. Applied on 35,527 labeled EBUS images, the CNN achieved 63.1% accuracy, with the highest F1-score in stations 4L, 4R, and 10R. Three expert bronchoscopists independently annotated Grad-CAM maps from 3131 test images. Activations predominantly aligned with lymph nodes and/or blood vessels, yielding an accuracy of 65.9% and an F1-score of 58.4%, with moderate interobserver agreement. These findings indicate that DL can aid lymph node station classification and that XAI offers meaningful insight into model behavior. The proposed framework may enhance anatomical orientation and operator training during EBUS, although further optimization and multicenter validation are required.
Background: Lung function (LF) development from adolescence to adulthood varies widely. Two related but distinct phenomena are relevant: pronounced growth (large positive change in FEV 1 z-score) and classical catch-up (crossing the lower limit of normal). We aimed to quantify both. Methods: We used data from 2169 participants in the HUNT Study with spirometry measured between 13 and 42 years. Change in FEV 1 z-score between baseline and follow-up was categorized into five FEV 1 change groups: pronounced increase, moderate increase, unchanged, moderate decline and pronounced decline. A catch-up group (z≤−1.645 to >−1.645) was also identified. Results: Of 82 individuals (3.8%) with pronounced LF increase from adolescence to adulthood, only 39 (48%) met the classical catch-up definition. Conversely, 48 of 87 (55%) individuals classified as catch-up did not show pronounced increase. Both groups shared a similar phenotype (female-predominant, low baseline BMI, and favorable adult metabolic profiles), but differed in adult LF: pronounced increase reached above-average FEV 1 (z=+0.65), while classical catch-up reached only partial normalization (z=−0.55). In contrast, the pronounced decline group had the highest BMI, C-reactive protein (CRP), and burden of chronic symptoms. Baseline characteristics differed between change groups, though individual distributions overlapped substantially. Conclusion: Pronounced LF growth and classical catch-up identify overlapping but distinct individuals with similar phenotypes yet different degrees of adult lung-function normalization. A substantial proportion of the most favorable lung-function gains originate from apparently normal baselines and are not detected by a binary catch-up definition.