Maximum signal intensity ratio on T1-weighted images for differentiating parotid gland tumors: a comparative study across two medical centers.

OBJECTIVE: To evaluate the value of the maximum signal intensity of tumor on T1-weighted images, using buccal subcutaneous fat as a reference, for differentiating parotid gland tumors across two medical centers. METHODS: T1-weighted images of 87 pleomorphic adenomas (PAs), 58 Warthin tumors (WTs), and 38 malignant tumors (MTs) from center A were retrospectively analyzed, compared with 45 PAs, 45 WTs, and 36 MTs from center B. The maximum signal intensity of tumor and the mean signal intensity of buccal subcutaneous fat were measured on T1-weighted images, then the tumor-to-fat signal intensity ratio (T1-max-SIR) was calculated. The distribution differences in T1-max-SIRs among the three groups of tumors within and between the two centers were statistically analyzed. Diagnostic performance was evaluated using receiver operating characteristic analysis. RESULTS: In both centers, T1-max-SIRs were significantly higher in WTs than in PAs (P < 0.001) and MTs (P < 0.001), while there was no statistically significant difference between PAs and MTs (center A: P = 0.257, center B: P = 0.543). In center A and center B, the AUC, sensitivity, and specificity for differentiating WTs from PAs were 0.862 vs. 0.833, 0.810 vs. 0.844, and 0.793 vs. 0.844, respectively, and the AUC, sensitivity, and specificity for differentiating WTs from MTs were 0.791 vs. 0.788, 0.810 vs. 0.844, and 0.711 vs. 0.778, respectively. CONCLUSIONS: T1-max-SIR is effective in differentiating WTs from PAs and MTs, demonstrating highly consistent diagnostic efficiency across the two medical centers. This method provides a valuable tool for the differential diagnosis of parotid gland tumors.