Use of extreme gradient boosting, light gradient boosting machine, and deep neural networks to evaluate the activity stage of extraocular muscles in thyroid-associated ophthalmopathy
Springer Science and Business Media LLC - Trang 1-8 - 2023
Tóm tắt
To develop a machine learning model to evaluate the activity stage of extraocular muscles in thyroid-associated ophthalmopathy (TAO). This study retrospectively analysed data from patients with TAO who underwent contrast-enhanced magnetic resonance imaging (MRI) from 2015 to 2022. Three independent machine learning models, namely, extreme gradient boosting (XGBoost), light gradient boosting machine (LightGBM), and deep neural networks (DNNs), were constructed using common clinical features. The performance of these models was compared using evaluation metrics such as the area under the receiver operating curve (AUC), accuracy, precision, recall, and F1 score. The importance of features was explained using Shapley additive explanations (SHAP). A total of 2561 eyes of 1479 TAO patients were included in this study. The original dataset was randomly divided into a training set (80%, n = 2048) and a test set (20%, n = 513). In the performance evaluation of the test set, the LightGBM model had the best diagnostic performance (AUC 0.9260). According to the SHAP results, features such as conjunctival congestion, swollen caruncles, oedema of the upper eyelid, course of TAO, and intraocular pressure had the most significant impact on the LightGBM model. This study used contrast-enhanced MRI as an objective evaluation criterion and constructed a LightGBM model based on readily accessible clinical data. The model had good classification performance, making it a promising artificial intelligence (AI)-assisted tool to help community hospitals evaluate the inflammatory activity of extraocular muscles in TAO patients in a timely manner.
Tài liệu tham khảo
Stein JD, Childers D, Gupta S et al (2015) Risk factors for developing thyroid-associated ophthalmopathy among individuals with Graves disease. JAMA Ophthalmol 133:290–296. https://doi.org/10.1001/jamaophthalmol.2014.5103
Kahaly GJ, Petrak F, Hardt J et al (2005) Psychosocial morbidity of Graves’ orbitopathy. Clin Endocrinol 63:395–402. https://doi.org/10.1111/j.1365-2265.2005.02352.x
Lee H, Roh HS, Yoon JS, Lee SY (2010) Assessment of quality of life and depression in Korean patients with Graves’ ophthalmopathy. Korean J Ophthalmol 24:65–72. https://doi.org/10.3341/kjo.2010.24.2.65
Bruscolini A, Sacchetti M, La Cava M et al (2018) Quality of life and neuropsychiatric disorders in patients with Graves’ orbitopathy: current concepts. Autoimmun Rev 17:639–43. https://doi.org/10.1016/j.autrev.2017.12.012
European Group of Graves’ Orbitopathy, Perros P, Baldeschi L et al (2006) A questionnaire survey on the management of Graves’ orbitopathy in Europe. Eur J Endocrinol 155:207–211. https://doi.org/10.1530/eje.1.02201
Dolman PJ (2012) Evaluating Graves’ orbitopathy. Best Pract Res Clin Endocrinol Metab 26:229–248. https://doi.org/10.1016/j.beem.2011.11.007
Bartalena L, Baldeschi L, Boboridis K et al (2016) The 2016 European Thyroid Association/European Group on Graves’ Orbitopathy Guidelines for the management of graves’ orbitopathy. Eur Thyroid J 5:9–26. https://doi.org/10.1159/000443828
Bartalena L, Kahaly GJ, Baldeschi L et al (2021) The 2021 European Group on Graves’ orbitopathy (EUGOGO) clinical practice guidelines for the medical management of Graves’ orbitopathy. Eur J Endocrinol 185:G43–G67. https://doi.org/10.1530/EJE-21-0479
Tachibana S, Murakami T, Noguchi H et al (2010) Orbital magnetic resonance imaging combined with clinical activity score can improve the sensitivity of detection of disease activity and prediction of response to immunosuppressive therapy for Graves’ ophthalmopathy. Endocr J 57:853–861. https://doi.org/10.1507/endocrj.k10e-156
Hiromatsu Y, Eguchi H, Tani J (2012) Management of Graves’ ophthalmopathy by using orbital magnetic resonance imaging. Nihon Rinsho 70:1932–1937
Siakallis LC, Uddin JM, Miszkiel KA (2018) Imaging investigation of thyroid eye disease. Ophthalmic Plast Reconstr Surg 34:S41–S51. https://doi.org/10.1097/IOP.0000000000001139
Tortora F, Cirillo M, Ferrara M et al (2013) Disease activity in Graves’ ophthalmopathy: Diagnosis with orbital MR imaging and correlation with clinical score. Neuroradiol J 26:555–564. https://doi.org/10.1177/197140091302600509
May M (2021) Eight ways machine learning is assisting medicine. Nat Med 27:2–3. https://doi.org/10.1038/s41591-020-01197-2
Liuv G, Xu J, Wang C et al (2022) A machine learning method for predicting the probability of MODS using only non-invasive parameters. Comput Methods Programs Biomed 227:107236. https://doi.org/10.1016/j.cmpb.2022.107236
Wang J, Chen H, Wang H et al (2023) A risk prediction model for physical restraints among older chinese adults in long-term care facilities: machine learning study. J Med Internet Res 25:e43815. https://doi.org/10.2196/43815
Hill NR, Ayoubkhani D, McEwan P et al (2019) Predicting atrial fibrillation in primary care using machine learning. PLoS One 14:e0224582. https://doi.org/10.1371/journal.pone.0224582
Sekelj S, Sandler B, Johnston E et al (2021) Detecting undiagnosed atrial fibrillation in UK primary care: validation of a machine learning prediction algorithm in a retrospective cohort study. Eur J Prevent Cardiol 28:598–605. https://doi.org/10.1177/2047487320942338
T. Chen, C. Guestrin (2016) Xgboost: A scalable tree boosting system. Proceedings of the 22nd ACM sigkdd international conference on knowledge discovery and data mining. 785–794. https://doi.org/10.1145/2939672.2939785
Kaur D, Uslu S, Rittichier KJ, Durresi A (2023) Trustworthy artificial intelligence: a review. https://doi.org/10.1145/3491209
Ke GL, Meng Q, Finley T et al (2017) LightGBM: a highly efficient gradient boosting decision tree. Adv Neural Inf Process Syst 30:52
Guyon I, Weston J, Barnhill S, Vapnik V (2002) Gene selection for cancer classification using support vector machines. Mach Learn 1–3:389–422. https://doi.org/10.1023/A:1012487302797
Fernandez A, Garcia S, Herrera F, Chawla NV (2018) SMOTE for learning from imbalanced data: progress and challenges. Marking 15-Year Anniversary J Artif Intell Res 16:341–378. https://doi.org/10.1613/jair.1.11192
LeCun Y, Bengio Y, Hinton G (2015) Deep learning. Nature 521:436–444. https://doi.org/10.1038/nature14539
Sze V, Chen YH, Yang TJ, Emer JS (2017) Efficient processing of deep neural networks: A tutorial and survey. Proc IEEE 105:2295–2329. https://doi.org/10.48550/arXiv.1703.09039
Lundberg SM, Lee SI (2017) A unified approach to interpreting model predictions. Adv Neural Inf Process Syst 30. https://doi.org/10.48550/arXiv.1705.07874
Ross EL, Zuromski KL, Reis BY et al (2021) Accuracy requirements for cost-effective suicide risk prediction among primary care patients in the US. JAMA Psychiat 78:642–650. https://doi.org/10.1001/jamapsychiatry.2021.0089
Laurberg P, Berman DC, Pedersen IB et al (2015) Double vision is a major manifestation in moderate to severe Graves’ orbitopathy, but it correlates negatively with inflammatory signs and proptosis. J Clin Endocrinol Metab 100:2098–2105. https://doi.org/10.1210/jc.2014-4557
Saeed P, Tavakoli Rad S, Bisschop P (2018) Dysthyroid opticneuropathy. Ophthalmic Plast Reconstr Surg 34:S60–S67. https://doi.org/10.1097/IOP.0000000000001146
Mourits MP, Koornneef L, Wiersinga WM et al (1989) Clinical criteria for the assessment of disease activity in Graves’ ophthalmopathy: a novel approach. Br J Ophthalmol 73:639–644. https://doi.org/10.1136/bjo.73.8.639