Objective To predict the stage I-II necrosis of the femoral head by a radiomics nomogram based on digital radiography (DR) of the hip joint to expand the application of routine DR in the evaluation of the stage I-II femoral head necrosis. Methods A total of 61 patients with Osteonecrosis of the femoral head (ONFH) and 24 normal healthy subjects were selected as the observation subjects. All the patients in this study received DR and MRI scans of the hip joint. We extracted 1409 radiomics features from all region of interests(ROIs) of the DR images. The minimum absolute contraction and selection operator (LASSO) regression method was used to select the features, and then two machine learning classification models Multilayer Perceptron (MLP) and Support Vector Machine (SVM) were constructed to detect femoral head necrosis. Incorporating radiomics signature(Rad-score) and independent demographics, the radiomics nomogram was established by logistic regression analysis. Receiver operating characteristics curve with area under the curve (AUC), sensitivity, specificity, and sensitivity were used to evaluate diagnostic performance. Results In the validation set, the AUC of the MLP and SVM radiomics models was 0.980 and 0.954, respectively, and the AUC of the radiomics nomogram was 0.981. Conclusion Machine learning based on X-ray radiomics features can assist in screening high-risk individuals with stages I-II femoral head necrosis, ultimately confirmed by MRI images.