Logistic regression-based model for differential diagnosis between plasma cell mastitis and non-mass type breast cancer.

Objective: To develop a diagnostic model for differentiating plasma cell mastitis (PCM) from non-mass type breast cancer (NMTBC) to facilitate clinical decision-making and reduce unnecessary invasive biopsies. Methods: A total of 120 patients with pathologically confirmed non-mass breast lesions (60 with PCM and 60 with NMTBC) were retrospectively included. Clinical and multimodal ultrasound features were assessed. Univariate analysis was performed to identify variables with differential diagnostic value. The dataset was randomly divided into a training set and a test set at a 7:3 ratio, and an elastic net regularized logistic regression model was constructed based on the training set. The diagnostic performance was validated in the independent test set, and the final model was visualized as a nomogram. Results: The regularized logistic regression model ultimately incorporated eight features: age, lesion intensity, location, calcification, resistive index (RI), spiculated margin, posterior acoustic features, and elastography score. In the testing set, the model achieved an area under the receiver operating characteristic curve (AUC) of 0.969 (95% CI: 0.921-1.000), with a sensitivity of 88.9% (95% CI: 67.2-96.9%), a specificity of 94.4% (95% CI: 74.2-99.0%), an accuracy of 91.7% (95% CI: 78.2-97.1%), a positive predictive value of 94.1% (95% CI: 73.0-99.0%), and a negative predictive value of 89.5% (95% CI: 68.6-97.1%). Conclusion: The proposed model effectively differentiates PCM from NMTBC, serving as a reliable clinical tool to help minimize unnecessary biopsies for benign lesions.