Assisting the diagnosis of Graves’ hyperthyroidism with Bayesian-type and SOM-type neural networks by making use of a set of three routine tests and their correlation with free T4

Development and preliminary validation of a machine learning system for thyroid dysfunction diagnosis based on routine laboratory tests

Background

Approximately 2.4 million patients in Japan would benefit from treatment for thyroid disease, including Graves’ disease and Hashimoto’s disease. However, only 450,000 of them are receiving treatment, and many patients with thyroid dysfunction remain largely overlooked. In this retrospective study, we aimed to develop and conduct preliminary testing on a machine learning method for screening patients with hyperthyroidism and hypothyroidism who would benefit from prompt medical treatment.

Methods

We collected electronic medical records and medical checkup data from four hospitals in Japan. We applied four machine learning algorithms to construct classification models to distinguish patients with hyperthyroidism and hypothyroidism from control subjects using routine laboratory tests. Performance evaluation metrics such as sensitivity, specificity, and the area under receiver operating characteristic (AUROC) were obtained. Techniques such as feature importance were further applied to understand the contribution of each feature to the machine learning output.

Results

The results of cross-validation and external evaluation indicated that we achieved high classification accuracies (AUROC = 93.8% for hyperthyroidism model and AUROC = 90.9% for hypothyroidism model). Serum creatinine (S-Cr), mean corpuscular volume (MCV), and total cholesterol were the three features that were most strongly correlated with the hyperthyroidism model, and S-Cr, lactic acid dehydrogenase (LDH), and total cholesterol were correlated with the hypothyroidism model.

Conclusions

We demonstrated the potential of machine learning approaches for diagnosing the presence of thyroid dysfunction from routine laboratory tests. Further validation, including prospective clinical studies, is necessary prior to application of our method in the clinic.

An overview of deep learning applications in precocious puberty and thyroid dysfunction

In the last decade, deep learning methods have garnered a great deal of attention in endocrinology research. In this article, we provide a summary of current deep learning applications in endocrine disorders caused by either precocious onset of adult hormone or abnormal amount of hormone production. To give access to the broader audience, we start with a gentle introduction to deep learning and its most commonly used architectures, and then we focus on the research trends of deep learning applications in thyroid dysfunction classification and precocious puberty diagnosis. We highlight the strengths and weaknesses of various approaches and discuss potential solutions to different challenges. We also go through the practical considerations useful for choosing (and building) the deep learning model, as well as for understanding the thought process behind different decisions made by these models. Finally, we give concluding remarks and future directions.

Assisting the diagnosis of overt hypothyroidism with pattern recognition methods, making use of a set of routine tests, and their multiple correlation with total T4

In our previous papers we proposed a novel screening method that assists the diagnosis of patients with overt Graves' hyperthyroidism by making use of routine test data and pattern recognition methods. This method can be applied by non-specialists during physical check-ups at low cost and is expected to lead to rapid referrals for examination and treatment by thyroid specialists, that is, to improve patients' QOL. In this report, we investigate whether a similar screening method is also applicable for overt hypothyroidism. Fifty-six subjects with 12 routine test data with a known diagnosis (30 patients with overt hypothyroidism and 52 healthy female volunteers, and 26 patients with overt hypothyroidism and 48 healthy male volunteers) were used as training data. Then, test samples of patients who had also undergone the same routine tests at the Tohoku university hospital were screened by our method for overt hypothyroidism. The present examination of the screening method showed its high screening ability with the set of four parameters used (lactate dehydrogenase [LDH], total cholesterol [TC], red blood cell [RBC] and serum creatinine [S-Cr]). It was found that there was a strong multiple correlation between the set of routine test parameters and serum total thyroxine (TT4) in the patients with overt hypothyroidism, which supports the usefulness of our screening method.

Assisting the diagnosis of Graves' hyperthyroidism with pattern recognition methods and a set of three routine tests parameters, and their correlations with free T4 levels: Extension to male patients

In our previous paper, we proposed a novel screening method that aids the diagnosis of female patients with Graves' hyperthyroidism via two types of neural networks and the use of routine test data. This method can be applied by non-specialists during physical checkups at a low cost and is expected to lead to rapid referrals for examination and treatment by thyroid specialists; i.e., to improve patients' QOL. In this report, we investigate whether the screening method is also applicable to males since sex differences exist in routine test data. The values of 14 routine test parameters for 78 subjects with definite diagnoses (31 patients with Graves' hyperthyroidism and 48 healthy volunteers) were adopted as training data, and 133 individuals who had also undergone the same routine tests at Tohoku University Hospital were screened for Graves' hyperthyroidism using our method. The present examination of our screening method in males showed its high screening ability with the set of parameters used (low serum creatinine, elevated alkaline phosphatase, and low total cholesterol). It was also found that there is strong multiple correlation between a set of three parameters and serum free thyroxine (FT4) in male patients with Graves' hyperthyroidism. A formula for FT4 consisting of three parameters was obtained, and this can be utilized in place of the true FT4 value. This result also supports the usefulness of our screening method.