Texture analysis on diffusion tensor imaging: discriminating glioblastoma from single brain metastasis

Area of Science:

• Radiology
• Oncology
• Medical Imaging

Background:

• Texture analysis is utilized in various radiological modalities for oncologic tumor staging and prognosis.
• Differentiating between glioblastoma (GBM) and single brain metastasis (MET) is crucial for treatment planning.

Purpose of the Study:

• To evaluate the diagnostic accuracy of magnetic resonance imaging texture analysis (MRTA) in distinguishing GBM from MET.
• To assess the heterogeneity of both the solid tumor and peritumoral edema using MRTA.

Main Methods:

• Preoperative 3-T MRI scans from 43 patients (22 GBM, 21 MET) were analyzed.
• MRTA was performed on diffusion tensor imaging (DTI) using TexRAD software, applying filtration histogram techniques.
• Heterogeneity was quantified using entropy, and receiver operating characteristic (ROC) analysis determined diagnostic thresholds.

Main Results:

• No significant difference in solid tumor heterogeneity was found between GBM and MET.
• Peritumoral edema heterogeneity was significantly higher in GBM compared to MET.
• This finding allowed differentiation with 80% sensitivity and 90% specificity.

Conclusions:

• Peritumoral edema heterogeneity assessed by MRTA enhances diagnostic accuracy in differentiating GBM from MET.
• This method can aid in medical staging and optimize surgical resection and postoperative management strategies.