Machine Learning Expands Across Endocrinology

Machine learning "might provide remarkable benefits to the endocrine field."

1

Machine learning applications in endocrinology have surged, particularly in thyroid-related research, with 1,130 studies identified from 2000 to 2024.
2

Thyroid disorders accounted for 68% of studies, focusing on ultrasound evaluations and demonstrating diagnostic performance comparable to expert radiologists.
3

ML models showed high accuracy in predicting malignancy and lymph node metastasis, with some achieving up to 100% accuracy in cytology analysis.
4

Limitations include lack of model transparency, small sample sizes, and infrequent external validation, highlighting a need for improved research quality.
5

Interdisciplinary collaboration is essential for integrating ML into clinical practice and realizing its potential benefits in endocrinology.

1