Objective To develop a diagnostic model for high-grade cervical squamous intraepithelial lesions to enhance diagnostic accuracy and inform clinical triage, contributing to reduced incidence of cervical cancer.

Methods Clinical data of 480 patients who underwent colposcopy and cervical biopsy between March 2022 and August 2023 at The Second Hospital of Jilin University were retrospectively analyzed. Using histopathological diagnosis as the gold standard, LASSO regression was used for feature screening and constructing a diagnostic nomogram. The patients were randomly assigned into training and validation sets in a 7:3 ratio for the model performance evaluation.

Results The model was screened for independent risk factors using LASSO regression and column line diagrams were constructed. Model validation showed excellent performance in diagnostic accuracy, with an area under the curve (AUC) of 0.9805 (95% confidence interval CI: 0.9699-0.9911) for the training set and 0.9629 (95% CI: 0.9375-0.9893) for the test set. Calibration curve analysis showed that the model predictions were highly consistent with the actual incidence, and clinical decision curve analysis showed that the model provided a noticeable net benefit at most thresholds, with good potential for clinical application.

Conclusions  The nomogram model constructed based on clinical data demonstrated a strong capability for diagnosing high-grade cervical squamous intraepithelial lesions.