Tumor-infiltrating lymphocytes evaluation using quantitative DCE-MRI-based habitat imaging in breast cancer patients: a preliminary study

Motivation: Tumor-infiltrating lymphocytes (TILs) are highly correlated with the prognosis of tumor treatment. We seek an efficient and noninvasive method for evaluating TILs based on habitat imaging of perfusion MRI. Goal(s): We investigate whether the quantitative DCE-based habitat imaging could be utilized to assess TILs in breast cancer patients. Approach: We developed a logistic predictive model utilizing extracted subregion DCE features to assess TILs, leveraging tumor spatial heterogeneity. Results: Significant perfusion differences exist between the low and medium-high TILs groups. The predictive model distinguishes the two groups with an AUC of 0.671, sensitivity of 0.571, and specificity of 0.75. Impact: This study indicates that quantitative DCE-based habitat imaging holds potential for the assessment of TILs in breast cancer.