BACKGROUND
Growth hormone deficiency (GHD) and idiopathic short stature (ISS) are the major etiologies of short stature in children; however, the current diagnostic approaches have various limitations.
OBJECTIVE
We aimed to develop a machine-learning-based model using sella magnetic resonance imaging (MRI)-based radiomics and clinical parameters to diagnose GHD and ISS.
METHODS
A total of 293 children with normal sella magnetic resonance imaging (MRI) findings in the training set and 47 children in the test set from different hospitals were enrolled. A total of 186 radiomic features were extracted from the pituitary glands using a semi-automatic segmentation process for both the T2-weighted and contrast-enhanced T1-image. The clinical parameters included auxological data, insulin-like growth factor-I (IGF-I), and bone age. The XGBoost algorithm was used to train the prediction models. Internal validation was conducted using five-fold cross-validation on the training set, and external validation was conducted on the test set. Model performance was assessed by plotting the area under the receiver operating characteristic curve (AUC). The mean absolute Shapley values were computed to quantify the impact of each parameter.
RESULTS
The AUCs of the clinical, radiomics, and combined models were 0.684, 0.691, and 0.830, respectively, in the external validation. Among the clinical parameters, the major contributing factors to prediction were body mass index standard deviation score (SDS), chronological age‒bone age, weight SDS, growth velocity, and IGF-I SDS in the clinical model. In the combined model, radiomic features added incremental value to the prediction (combined model vs. clinical model, p = 0.034; combined model vs. radiomics model, p = 0.019).
CONCLUSIONS
These findings highlight the potential of machine learning-based models using radiomics and clinical parameters for diagnosing GHD and ISS.