Nomogram Model Construction and Validation with Virtual Touch Tissue Imaging Quantification and Clinicopathological Features to Predict Recurrence in Breast Cancer

Background: This study aimed to develop a nomogram model integrating virtual touch tissue imaging quantification (VTIQ) with clinicopathological features to predict postoperative breast cancer recurrence, guide individualized treatment, and improve prognosis. Methods: This study retrospectively included 420 female patients who underwent radical mastectomy for breast cancer and received an elastography touch imaging quantification examination before surgery at our hospital (2017–2022). The patients were divided into training and validation sets at a ratio of 7:3. After a 3-year follow-up, both cohorts were stratified into recurrence and non-recurrence groups. Clinicopathologic characteristics and VTIQ parameters (shear wave velocity, SWV) were compared between the two groups. A nomogram for predicting postoperative recurrence in breast cancer was developed using multivariable logistic regression. The performance was evaluated using a receiver operating characteristic (ROC) analysis, calibration assessment, and decision curve analysis (DCA) to assess discrimination, calibration, and clinical usefulness. Results: The training set showed significantly higher SWV values in recurrent patients than in non-recurrent patients (p < 0.05). Logistic regression identified histological grade (odds ratio (OR): 3.36, 95% confidence interval (CI): 1.23–9.19), calcification (OR: 3.16, 95% CI: 1.15–8.68), estrogen receptor (ER)/progesterone receptor (PR) (OR: 2.74, 95% CI: 1.03–7.31), and SWV (OR: 3.71, 95% CI: 1.75–7.84) as independent predictive factors for postoperative recurrence of breast cancer (p < 0.05). The area under the ROC curve (AUROC) was 0.789 (95% CI: 0.729–0.850) for the training set and 0.728 (95% CI: 0.615–0.841) for the validation set. These findings indicate that the nomogram model demonstrates good discrimination for the postoperative recurrence of breast cancer. Calibration and DCA curves confirmed that the predicted probabilities of the model closely matched the actual pathological grading results, demonstrating the clinical utility of the model. Conclusions: The nomogram model integrating VTIQ parameters with clinicopathological features demonstrates good predictive value for postoperative recurrence of breast cancer. This model provides an important reference for identifying patients at high risk of recurrence before surgery and may improve patient prognosis.