Performance of low- and high-temporal-resolution DCE-MRI texture analysis in distinguishing breast lesions from background enhancement.

OBJECTIVES

To investigate the diagnostic potential of texture-based analysis of dynamic contrast-enhanced MRI (DCE-MRI) for breast lesions and background enhancement (BE).

METHODS

This retrospective study analyzed 62 patients who underwent preoperative high-temporal resolution DCE-MRI (1+26 phases), including 39 malignant and 23 benign lesions. A control group of 78 patients received preoperative low-temporal resolution DCE-MRI (1+5 phases), comprising 46 malignant and 32 benign lesions. All patients also underwent conventional T1WI, T2WI MRI scans, and DCE-MRI. Quantitative parameters were obtained using a two-compartment Extended Tofts model, calculating pharmacokinetic parameters: volume transfer constant (K), rate constant (K), extravascular extracellular volume fraction (V), and fractional plasma volume (V). Texture features based on the K map were extracted. The region of interest for the lesion center, surrounding peripheral area, and BE was delineated. Receiver operating characteristic (ROC) analysis was used to evaluate the diagnostic performance of the K texture features model.

RESULTS

Pharmacokinetic parameters significantly differed between high-temporal resolution and low-temporal resolution DCE-MRI (P < 0.05). In the malignant group, the average K of the lesion area from high-temporal resolution DCE-MRI was significantly correlated with pathological grading (r = 0.400, P = 0.012). There were significant differences in the mean values of K, K, V, V and time to peak (TTP) between the two DCE-MRI groups across the lesion, peri-lesional, and BE areas. In the differentiation between benign and malignant lesions, ROC analysis demonstrated that high-temporal resolution DCE-MRI provided slight but significant advantages in differentiating benign and malignant lesions in the lesion center, BE areas.

CONCLUSIONS

Texture analysis based on high-temporal resolution DCE-MRI may potentially improve breast cancer diagnostic performance. Specifically, combining the lesion, BE area, and K-mean parameters contributes to the diagnosis of breast lesions, background enhancement, and the pathological grading of malignant tumors.