Treffer: Evaluating the efficacy of the ResNet50 deep learning model utilizing thyroid scintigraphy images for predicting the outcomes of initial iodine-131 therapy in patients with Graves' disease.
Original Publication: London : Chapman and Hall in association with the British Nuclear Medicine Society, c1980-
Reid JR, Wheeler SF. Hyperthyroidism: diagnosis and treatment. Am Fam Physician 2005; 72:623–630.
Rivkees SA, Dinauer C. An optimal treatment for pediatric Graves’ disease is radioiodine. J Clin Endocrinol Metab 2007; 92:797–800.
Ylli D, Van Nostrand D, Wartofsky L. Conventional radioiodine therapy for differentiated thyroid cancer. Endocrinol Metab Clin North Am 2019; 48:181–197.
Cappelli C, Braga M, De Martino E, Castellano M, Gandossi E, Agosti B, et al. Outcome of patients surgically treated for various forms of hyperthyroidism with differentiated thyroid cancer: experience at an endocrine center in Italy. Surg Today 2006; 36:125–130.
Litjens G, Kooi T, Bejnordi BE, Setio AAA, Ciompi F, Ghafoorian M, et al. A survey on deep learning in medical image analysis. Med Image Anal 2017; 42:60–88.
Giovanella L, Avram AM, Ovčariček PP, Clerc J. Thyroid functional and molecular imaging. Presse Med 2022; 51:104116.
Duan L, Zhang H-Y, Lv M, Zhang H, Chen Y, Wang T, et al. Machine learning identifies baseline clinical features that predict early hypothyroidism in patients with Graves’ disease after radioiodine therapy. Endocr Connect 2022; 11:e220119.
De Leo S, Lee SY, Braverman LE. Hyperthyroidism. Lancet 2016; 388:906–918.
Jonklaas J, Bianco AC, Bauer AJ, Burman KD, Cappola AR, Celi FS, et al.; American Thyroid Association Task Force on Thyroid Hormone Replacement. Guidelines for the treatment of hypothyroidism: prepared by the American Thyroid Association task force on thyroid hormone replacement. Thyroid 2014; 24:1670–1751.
Ji L, Mao R, Wu J, Ge C, Xiao F, Xu X, et al. Deep convolutional neural network for nasopharyngeal carcinoma discrimination on MRI by comparison of hierarchical and simple layered convolutional neural networks. Diagnostics (Basel) 2022; 12:2478.
Ahn JH, Kim HC, Rhim JK, Park JJ, Sigmund D, Park MC, et al. Multi-view convolutional neural networks in rupture risk assessment of small, unruptured intracranial aneurysms. J Pers Med 2021; 11:239.
Garga B, Abboubakar H, Sourpele RS, Gwet DLL, Bitjoka L. Pollen grain classification using some convolutional neural network architectures. J Imaging 2024; 10:158.
Shi G-Y, Wu H-P, Luo S-H, Lu X-Y, Ren B, Zhang Q, et al. 1D gradient-weighted class activation mapping, visualizing decision process of convolutional neural network-based models in spectroscopy analysis. Anal Chem 2023; 95:9959–9966.
Chen W-C, Liu P-Y, Lai C-C, Lin Y-H. Identification of environmental microorganism using optimally fine-tuned convolutional neural network. Environ Res 2022; 206:112610.
Oya M, Sugimoto S, Sasai K, Yokoyama K. Investigation of clinical target volume segmentation for whole breast irradiation using three-dimensional convolutional neural networks with gradient-weighted class activation mapping. Radiol Phys Technol 2021; 14:238–247.
Sohn E. The reproducibility issues that haunt health-care AI. Nature 2023; 613:402–403.
Lenharo M. An AI revolution is brewing in medicine. What will it look like? Nature 2023; 622:686–688.
Guo Y, Xu J, Li X, Zheng L, Pan W, Qiu M, et al. Classification and diagnosis of residual thyroid tissue in SPECT images based on fine-tuning deep convolutional neural network. Front Oncol 2021; 11:762643.
Peng S, Liu Y, Lv W, Liu L, Zhou Q, Yang H, et al. Deep learning-based artificial intelligence model to assist thyroid nodule diagnosis and management: a multicentre diagnostic study. Lancet Digit Health 2021; 3:e250–e259.
Dov D, Elliott Range D, Cohen J, Bell J, Rocke DJ, Kahmke RR, et al. Deep-learning-based screening and ancillary testing for thyroid cytopathology. Am J Pathol 2023; 193:1185–1194.
Chen L, Chen H, Pan Z, Xu S, Lai G, Chen S, et al. ThyroidNet: a deep learning network for localization and classification of thyroid nodules. Comput Model Eng Sci 2023; 139:361–382.
Nojima S, Kadoi T, Suzuki A, Kato C, Ishida S, Kido K, et al. Deep learning-based differential diagnosis of follicular thyroid tumors using histopathological images. Mod Pathol 2023; 36:100296.
Xu W, Fu Y-L, Zhu D. ResNet and its application to medical image processing: research progress and challenges. Comput Methods Programs Biomed 2023; 240:107660.
Zhou L-Q, Zeng S-E, Xu J-W, Lv W-Z, Mei D, Tu J-J, et al. Deep learning predicts cervical lymph node metastasis in clinically node-negative papillary thyroid carcinoma. Insights Imaging 2023; 14:222.
Ren W, Zhu Y, Wang Q, Song Y, Fan Z, Bai Y, Lin D. Deep learning prediction model for central lymph node metastasis in papillary thyroid microcarcinoma based on cytology. Cancer Sci 2023; 114:4114–4124.
Sarigül M, Ozyildirim BM, Avci M. Differential convolutional neural network. Neural Netw 2019; 116:279–287.
0 (Iodine-131)
Weitere Informationen
Background: Graves' disease (GD), a leading cause of hyperthyroidism, exhibits heterogeneous responses to iodine-131 therapy, underscoring the need for accurate predictive tools. While pertechnetate ( 99m TcO 4- ) thyroid scintigraphy provides essential functional insights, its clinical utility is restricted by limited resolution and the subjective interpretation of results. Deep learning emerges as a promising approach for extracting prognostic features from imaging data.
Objective: This research seeks to construct a radiomic model employing deep learning methodologies, utilizing thyroid scintigraphy, to predict the efficacy of initial iodine-131 therapy in patients diagnosed with GD.
Methods: This retrospective study involved 121 patients diagnosed with GD who underwent pretreatment thyroid scintigraphy. A ResNet50 convolutional neural network was implemented, trained end-to-end on standardized thyroid scintigraphy data to classify treatment outcomes as either cured or not cured. Gradient-weighted class activation mapping (Grad-CAM) was employed to identify region-specific predictive features. Radiomic features extracted from deep learning were further analyzed using least absolute shrinkage and selection operator regression, with a random forest (RF) model serving as the final predictive tool. Performance metrics, including the area under the receiver operating characteristic curve (AUC), sensitivity, and specificity, were evaluated on independent training and validation datasets, using an 8 : 2 split.
Results: The ResNet50 model yielded AUC values of 0.828 [95% confidence interval (CI): 0.740-0.915] and 0.806 (95% CI: 0.656-0.956) for the training and validation datasets, respectively. A set of 16 critical radiomic features, extracted from the global pooling layer of ResNet50, enabled the construction of an RF classifier that exhibited enhanced generalization performance, as evidenced by a test AUC of 0.924. Decision curve analysis substantiated the model's clinical utility across a broad spectrum of threshold probabilities (10-90%). Grad-CAM visualizations revealed that the model predominantly concentrated on regions of thyroid parenchyma associated with treatment-responsive tissue.
Conclusion: This study introduces a pioneering deep learning framework that employs thyroid scintigraphy radiomics to predict iodine-131 therapeutic outcomes in GD. By combining ResNet50 for feature extraction with interpretable Grad-CAM localization, this approach advances personalized nuclear medicine strategies and mitigates the 'black box' limitation typically associated with artificial intelligence models. These findings require validation across multicenter cohorts to refine and optimize precision treatment protocols.
(Copyright © 2025 The Author(s). Published by Wolters Kluwer Health, Inc.)