{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,27]],"date-time":"2026-10-27T20:23:15Z","timestamp":1793132595072,"version":"4.5.0"},"reference-count":55,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2025,12,5]],"date-time":"2025-12-05T00:00:00Z","timestamp":1764892800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Endocrinol."],"abstract":"<jats:sec>\n                    <jats:title>Background<\/jats:title>\n                    <jats:p>Thyroid disease is a common endocrine disorder, with the differentiation between benign and malignant nodules being critical for clinical decision-making. Traditional diagnostic methods, such as ultrasound and TI-RADS classification, are limited by interobserver variability and time-consuming processes. While deep learning approaches such as CNNs and transformers have shown promise, they face challenges in multiscale feature extraction, global dependency modeling, and alignment with clinical standards.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>We proposes THMSNet, a hybrid architecture that integrates a pyramid structure for multiscale feature extraction and Mamba for global long-range dependency modeling. The serial channel\u2013spatial attention module (SCSAM) enhances feature representation, whereas the truth\u2013value calibration (TVC) algorithm aligns model predictions with pathological standards. The system is evaluated on a public dataset of 7,288 thyroid ultrasound images (3,282 benign, 4,006 malignant) via five metrics: accuracy, precision, recall, F1 score, and AUROC.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>THMSNet achieves 91.15% accuracy, 93.28% recall, and 96.92% AUROC, outperforming ResNet (86.03% accuracy) and DenseNet (95.50% AUROC). Confidence intervals are calculated for key metrics, further strengthening the rigor of results. Ablation studies confirm the utility of each module, with the pyramid architecture (+7.83% accuracy), Mamba (+2.99%), SCSAM (+6.94%), and TVC (+6.94%) progressively contributing to performance improvements.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusion<\/jats:title>\n                    <jats:p>THMSNet provides a robust and clinically applicable solution for thyroid nodule diagnosis, combining advanced feature extraction, attention mechanisms, and probability calibration. Its high accuracy and interpretability make it a valuable tool for assisting radiologists in clinical practice.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.3389\/fendo.2025.1686248","type":"journal-article","created":{"date-parts":[[2025,12,5]],"date-time":"2025-12-05T05:13:25Z","timestamp":1764911605000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":3,"title":["Thyroid intelligent diagnosis based on THMSNet"],"prefix":"10.3389","volume":"16","author":[{"given":"Zhen","family":"Rao","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Tao","family":"Yu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xitan","family":"Yu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2025,12,5]]},"reference":[{"key":"B1","doi-asserted-by":"crossref","DOI":"10.1177\/10507256251363120","article-title":"2025 American Thyroid Association management guidelines for adult patients with differentiated thyroid cancer","volume":"35","author":"Ringel","year":"2025","journal-title":"Thyroid"},{"key":"B2","doi-asserted-by":"publisher","DOI":"10.1016\/j.jacr.2017.01.046","article-title":"ACR thyroid imaging, reporting and data system (TI-RADS): white paper of the ACR TI-RADS committee","volume":"14","author":"Tessler","year":"2013","journal-title":"J Am Coll Radiol"},{"key":"B3","doi-asserted-by":"publisher","DOI":"10.1016\/j.jacr.2015.07.011","article-title":"Thyroid ultrasound reporting lexicon: White paper of the ACR thyroid imaging, reporting and data system (TIRADS) committee","volume":"12","author":"Grant","year":"2017","journal-title":"J Am Coll Radiol"},{"key":"B4","doi-asserted-by":"crossref","first-page":"730","DOI":"10.1007\/s12020-024-03731-5","article-title":"Interobserver variability in thyroid ultrasound","volume":"85","author":"de Carlos","year":"2024","journal-title":"Endocrine"},{"key":"B5","doi-asserted-by":"crossref","first-page":"4641","DOI":"10.21037\/qims-24-1796","article-title":"Development and validation of a nomogram based on conventional and contrast-enhanced ultrasound for differentiating malignant from benign thyroid nodules","volume":"15","author":"Wang","year":"2025","journal-title":"Quantitative Imaging in Medicine and Surgery"},{"key":"B6","article-title":"Correct and incorrect recommendations for or against fine needle biopsies of hypofunctioning thyroid nodules: performance of different ultrasound-based risk stratification systems","volume":"63","author":"Petersen","year":"2024","journal-title":"Nuklearmedizin-NuclearMedicine"},{"key":"B7","doi-asserted-by":"crossref","DOI":"10.1177\/00368504251346906","article-title":"Diagnostic performance of the ultrasound-based artificial intelligence diagnostic system in predicting cervical lymph node metastasis in patients with thyroid cancer: A systematic review and meta-analysis","volume":"108","author":"Tang","year":"2025","journal-title":"Science Progress"},{"key":"B8","article-title":"Building radiomics models based on ACR TI-RADS combining clinical features for discriminating benign and malignant thyroid nodules","volume":"16","author":"Chen","year":"2025","journal-title":"Front Endocrinol"},{"key":"B9","doi-asserted-by":"crossref","first-page":"544","DOI":"10.1186\/s12885-025-13917-3","article-title":"Thyroid nodule classification in ultrasound imaging using deep transfer learning","volume":"25","author":"Xu","year":"2025","journal-title":"BMC Cancer"},{"key":"B10","first-page":"1","article-title":"Thyroid cancer classification using transfer learning[C]\/\/2021 international conference on computer science and engineering (IC2SE)","volume-title":"IEEE","author":"Nugroho","year":"2021"},{"key":"B11","doi-asserted-by":"publisher","first-page":"4214","DOI":"10.1038\/s41598-025-86819-w","article-title":"ThyroNet-X4 genesis: An advanced deep learning model for auxiliary diagnosis of thyroid nodules\u2019 Malignancy","volume":"15","author":"Wang","year":"2025","journal-title":"Sci Rep"},{"key":"B12","doi-asserted-by":"crossref","DOI":"10.1038\/s41598-025-24651-y","article-title":"Automated thyroid nodule classification in ultrasound imaging using a hybrid vision transformer and Wasserstein GAN with gradient penalty","volume":"15","author":"Sujini","year":"2025","journal-title":"Scientific Reports"},{"key":"B13","doi-asserted-by":"publisher","first-page":"104101","DOI":"10.1016\/j.medengphy.2024.104101","article-title":"SRT: Swin-residual transformer for benign and Malignant nodules classification in thyroid ultrasound images","volume":"124","author":"Huang","year":"2024","journal-title":"Med Eng Phys"},{"key":"B14","article-title":"Thyroid Nodule Ultrasound Image Segmentation Based on Improved Swin Transformer","author":"Wu","year":"2025","journal-title":"IEEE Access"},{"key":"B15","article-title":"Enhancing thyroid nodule assessment with UTV-ST swin kansformer: a multimodal approach to predict invasiveness","author":"Zhao","year":"2025","journal-title":"IEEE Access"},{"key":"B16","article-title":"A Multi-Scale Model Based on Squeeze and Excitation Network for Classifying Obstacles in Front of Vehicles in Autonomous Driving","author":"Liu","year":"2025","journal-title":"IEEE Internet of Things Journal"},{"key":"B17","article-title":"Res4net-CBAM: A deep cnn with convolution block attention module for tea leaf disease diagnosis","volume":"83","author":"Bhuyan","year":"2024","journal-title":"Multimedia Tools and Applications"},{"key":"B18","doi-asserted-by":"crossref","DOI":"10.1016\/j.fraope.2024.100170","article-title":"Squeeze & excitation joint with combined channel and spatial attention for pathology image super-resolution","volume":"8","author":"Hayat","year":"2024","journal-title":"Franklin Open"},{"key":"B19","doi-asserted-by":"crossref","DOI":"10.1109\/ACCESS.2024.3460405","article-title":"Coordinate attention enhanced adaptive spatiotemporal convolutional networks for traffic flow forecasting","author":"Wei","year":"2024","journal-title":"IEEE Access"},{"key":"B20","first-page":"05561","article-title":"Global attention mechanism: Retain information to enhance channel-spatial interactions","volume":"2112","author":"Liu","year":"2021","journal-title":"ArXiv"},{"key":"B21","doi-asserted-by":"publisher","DOI":"10.1016\/j.heliyon.2024.e24188","article-title":"Leveraging Bayesian deep learning and ensemble methods for uncertainty quantification in image classification: A ranking-based approach","volume":"10","author":"Abdullah","year":"2024","journal-title":"Heliyon"},{"key":"B22","doi-asserted-by":"publisher","DOI":"10.12659\/MSM.926096","article-title":"Ensemble deep learning model for multicenter classification of thyroid nodules on ultrasound images","volume":"26","author":"Wei","year":"2020","journal-title":"Med Sci Monitor"},{"key":"B23","doi-asserted-by":"crossref","DOI":"10.1109\/ACCESS.2025.3552934","article-title":"Conformal prediction for uncertainty quantification and reliable HER2 status classification in breast cancer IHC images","author":"Pintawong","year":"2025","journal-title":"IEEE Access"},{"key":"B24","first-page":"1","article-title":"Tiny Deep Ensemble: Uncertainty Estimation in Edge AI Accelerators via Ensembling Normalization Layers with Shared Weights","author":"Ahmed","year":"2024"},{"key":"B25","doi-asserted-by":"crossref","first-page":"5186","DOI":"10.3390\/app15095186","article-title":"A systematic analysis of neural networks, fuzzy logic and genetic algorithms in tumor classification","volume":"15","author":"Al-Ashoor","year":"2025","journal-title":"Applied Sciences"},{"key":"B26","doi-asserted-by":"crossref","DOI":"10.1109\/CVPR.2017.106","article-title":"Feature pyramid networks for object detection","author":"Lin","year":"2017"},{"key":"B27","article-title":"Mamba: Linear-time sequence modeling with selective state spaces","author":"Gu","year":"2023","journal-title":"arXiv preprint arXiv:2312.00752"},{"key":"B28","doi-asserted-by":"crossref","DOI":"10.1109\/ICCV48922.2021.00009","article-title":"CvT: Introducing convolutions to vision transformers","author":"Wu","year":"2021"},{"key":"B29","doi-asserted-by":"crossref","DOI":"10.1109\/CVPR52688.2022.01167","article-title":"A ConvNet for the 2020s","author":"Liu","year":"2022"},{"key":"B30","doi-asserted-by":"crossref","DOI":"10.1109\/CVPR.2017.243","article-title":"Densely connected convolutional networks","author":"Huang","year":"2017"},{"key":"B31","first-page":"1","article-title":"EnhanceNet: Rethinking Model Scaling for Convolutional Neural Network","author":"Kolhe","year":"2024"},{"key":"B32","doi-asserted-by":"crossref","first-page":"109542","DOI":"10.1016\/j.compbiomed.2024.109542","article-title":"Advancing brain tumor classification: A robust framework using EfficientNetV2 transfer learning and statistical analysis","volume":"185","author":"Hassan","year":"2025","journal-title":"Computers in Biology and Medicine"},{"key":"B33","doi-asserted-by":"crossref","DOI":"10.1109\/CVPR42600.2020.00165","article-title":"Ghostnet: More features from cheap operations","author":"Han","year":"2020"},{"key":"B34","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1704.04861","article-title":"MobileNets: Efficient convolutional neural networks for mobile vision applications","author":"Howard","year":"2017","journal-title":"arXiv preprint arXiv.1704.04861"},{"key":"B35","doi-asserted-by":"crossref","DOI":"10.1109\/ICCV48922.2021.00061","article-title":"Pyramid vision transformer: A versatile backbone for dense prediction without convolutions","author":"Wang","year":"2021"},{"key":"B36","doi-asserted-by":"crossref","DOI":"10.1109\/CVPR42600.2020.01044","article-title":"Designing network design spaces","author":"Radosavovic","year":"2020"},{"key":"B37","doi-asserted-by":"crossref","DOI":"10.1109\/CVPR.2016.90","article-title":"Deep residual learning for image recognition","author":"He","year":"2016"},{"key":"B38","doi-asserted-by":"crossref","DOI":"10.1109\/ICCV48922.2021.00986","article-title":"Swin transformer: Hierarchical vision transformer using shifted windows","author":"Liu","year":"2021"},{"key":"B39","doi-asserted-by":"crossref","DOI":"10.1109\/ICCV48922.2021.00060","article-title":"Tokens-to-Token ViT: Training vision transformers from scratch on ImageNet","author":"Yuan","year":"2021"},{"key":"B40","article-title":"Very deep convolutional networks for large-scale image recognition","author":"Simonyan","year":"2014","journal-title":"arXiv preprint arXiv:1409.1556"},{"key":"B41","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2010.11929","article-title":"An image is worth 16x16 words: Transformers for image recognition at scale","author":"Dosovitskiy","year":"2020","journal-title":"arXiv preprint arXiv:2010.11929"},{"key":"B42","doi-asserted-by":"crossref","DOI":"10.1109\/CVPR.2018.00716","article-title":"ShuffleNet: An extremely efficient convolutional neural network for nobile devices","author":"Zhang","year":"2018"},{"key":"B43","doi-asserted-by":"publisher","first-page":"52010","DOI":"10.1109\/ACCESS.2020.2980290","article-title":"Comparison study of radiomics and deep learning based methods for thyroid nodules classification using ultrasound images","volume":"8","author":"Wang","year":"2020","journal-title":"IEEE Access"},{"key":"B44","doi-asserted-by":"publisher","DOI":"10.1002\/mp.14543","article-title":"Thyroid nodules risk stratification through deep learning based on ultrasound images","volume":"47","author":"Bai","year":"2020","journal-title":"Medical Physics"},{"key":"B45","doi-asserted-by":"publisher","DOI":"10.1089\/thy.2019.0752","article-title":"AIBx, artificial intelligence model to risk stratify thyroid nodules","volume":"30","author":"Thomas","year":"2020","journal-title":"Thyroid"},{"key":"B46","doi-asserted-by":"publisher","first-page":"221","DOI":"10.1016\/j.ultras.2016.09.011","article-title":"A pre-trained convolutional neural network based method for thyroid nodule diagnosis","volume":"73","author":"Ma","year":"2016","journal-title":"Ultrasonics"},{"key":"B47","doi-asserted-by":"crossref","DOI":"10.1002\/mp.12134","article-title":"Cascade convolutional neural networks for automatic detection of thyroid nodules in ultrasound images","volume":"44","author":"Ma","year":"2017","journal-title":"Medical Physics"},{"key":"B48","article-title":"Thyroid nodule benignty prediction by deep feature extraction. Bioinformatics and bioengineering","author":"Mei","year":"2017"},{"key":"B49","doi-asserted-by":"crossref","first-page":"108100","DOI":"10.1016\/j.cmpb.2024.108100","article-title":"Deep neural networks can differentiate thyroid pathologies on infrared hyperspectral images","volume":"247","author":"Baffa","year":"2024","journal-title":"Computer Methods and Programs in Biomedicine"},{"key":"B50","doi-asserted-by":"crossref","first-page":"544","DOI":"10.1186\/s12885-025-13917-3","article-title":"Thyroid nodule classification in ultrasound imaging using deep transfer learning","volume":"25","author":"Xu","year":"2025","journal-title":"BMC Cancer"},{"key":"B51","article-title":"Multiscale patch-based image restoration","volume":"25","author":"Papyan","year":"2015","journal-title":"IEEE Trans image Process"},{"key":"B52","article-title":"A comprehensive survey on source-free domain adaptation","volume":"46","author":"Li","year":"2024","journal-title":"IEEE Transactions on Pattern Analysis and Machine Intelligence"},{"key":"B53","doi-asserted-by":"crossref","first-page":"1750","DOI":"10.3390\/electronics14091750","article-title":"Federated learning in smart healthcare: A survey of applications, challenges, and future directions","volume":"14","author":"Nasajpour","year":"2025","journal-title":"Electronics"},{"key":"B54","first-page":"1","article-title":"A survey on Bayesian deep learning","volume":"53","author":"Wang","year":"2020","journal-title":"ACM Comput Surveys (csur)"},{"key":"B55","doi-asserted-by":"crossref","first-page":"123052","DOI":"10.1016\/j.eswa.2023.123052","article-title":"Deep semi-supervised learning for medical image segmentation: A review","volume":"245","author":"Han","year":"2024","journal-title":"Expert Systems with Applications"}],"container-title":["Frontiers in Endocrinology"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fendo.2025.1686248\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,12,5]],"date-time":"2025-12-05T05:13:30Z","timestamp":1764911610000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fendo.2025.1686248\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,12,5]]},"references-count":55,"alternative-id":["10.3389\/fendo.2025.1686248"],"URL":"https:\/\/doi.org\/10.3389\/fendo.2025.1686248","relation":{},"ISSN":["1664-2392"],"issn-type":[{"value":"1664-2392","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,12,5]]},"article-number":"1686248"}}