Perilesional Diffusion Tensor Imaging for Differentiating Malignant and Benign Causes of Intracerebral Hemorrhage

Background/Objectives: To evaluate the discriminative value of perilesional microstructural alterations assessed by diffusion tensor imaging (DTI) parameters in differentiating malignant and benign causes of intracerebral hemorrhage. Methods: This retrospective study included patients with intracerebral hemorrhage classified as benign or malignant based on follow-up findings and histopathological diagnosis. Fractional anisotropy (FA), axial diffusivity (AD), radial diffusivity (RD), mean diffusivity (MD), lesion-to-normal (L/N) ratios, and perilesional edema index (PEI) were extracted from standardized perilesional regions. Diffusion parameters were compared between groups, and their discriminative performance was assessed using ROC analysis and logistic regression. Results: A total of 41 patients (20 benign, 21 malignant) were included. Malignant cases demonstrated significantly greater PEI and distinct diffusion abnormalities, with markedly higher AD, RD, and MD values, as well as elevated L/N ratios (all p < 0.001). FA was significantly reduced in malignant cases (p = 0.016). Conclusions: Diffusivity metrics-particularly AD, RD, and MD values and their L/N ratios-demonstrated excellent diagnostic performance in differentiating malignant from benign causes of intracerebral hemorrhage, further informed by the PEI, and may support clinical decision-making in diagnostically challenging cases.