CT Image-Based Texture Analysis to Predict Microvascular Invasion in Primary Hepatocellular Carcinoma

Area of Science:

• Radiology
• Oncology
• Medical Imaging

Background:

• Microvascular invasion (MVI) is a critical prognostic factor in primary hepatocellular carcinoma (HCC).
• Accurate preoperative prediction of MVI is essential for guiding treatment strategies and improving patient outcomes.
• Current methods for MVI detection often require invasive procedures or lack sufficient predictive accuracy.

Purpose of the Study:

• To evaluate the clinical utility of computed tomography (CT) image-based texture analysis for predicting MVI in primary HCC.
• To identify optimal texture parameters from different CT phases that can differentiate MVI-positive from MVI-negative HCC.
• To establish a non-invasive approach for MVI prediction in HCC patients.

Main Methods:

• Retrospective analysis of CT images from 102 HCC patients (34 MVI-negative, 68 MVI-positive).
• Extraction of texture features from pre-contrast, arterial, portal, and equilibrium phase CT images using MaZda software.
• Selection of optimal texture parameters using Fisher's coefficient, POE+ACC, and MI, followed by receiver operating characteristic (ROC) curve analysis.

Main Results:

• Significant differences in Edmonson-Steiner grades, tumor size, margin, and intratumoral arteries between MVI-negative and MVI-positive groups.
• 58 texture parameters showed significant differences between the groups (P < 0.001).
• Optimal texture parameters from different phases achieved AUCs ranging from 0.754 to 0.766, with gray-level run-length matrix parameters being most valuable.

Conclusions:

• CT image-based texture analysis serves as a promising non-invasive tool for predicting MVI in primary HCC.
• Texture analysis can aid in preoperative risk stratification, treatment guidance, and prognosis evaluation for HCC patients.
• Further validation of these findings in larger cohorts is warranted to integrate texture analysis into routine clinical practice.