Texture Analysis of Histology Images for Characterizing Ultrasound-Stimulated Microbubble Radiation Enhancement Treatment Response.

Ultrasound-stimulated microbubble (USMB) therapy, in combination with radiotherapy (XRT), represents a promising approach to enhancing the efficacy of conventional cancer treatments by targeting tumor vasculature. Recent preclinical studies using MRI-guided focused ultrasound have demonstrated that USMB enhances radiation effects in tumor blood vessels, resulting in significantly greater tumor cell death than radiation alone. Dynamic contrast-enhanced MRI (DCE-MRI) has been instrumental in this methodology in mapping tumor perfusion heterogeneity, allowing for precise targeting of additional USMB and XRT to specific vascular regions. This study employed four advanced texture analysis methods, GLCM, GLDM, GLSZM, and NGTDM, to quantitatively assess changes in the cellular structure of prostate tumors following different treatments, including combinations of USMB and XRT targeted to low- and high-perfusion regions. Texture features, particularly those derived from GLCM, GLDM, and GLSZM, revealed significant differences in cell structure patterns across treatment groups. The GLSZM methodology was identified as the most sensitive method for detecting treatment-induced structural changes, effectively identifying regions of necrosis and varied stages of cell death. Texture-derivative analyses further highlighted intra-tumoral heterogeneity, especially in response to additional USMB + XRT treatments. These results align with findings in other tissue models, underscoring the value of texture analysis for monitoring treatment response.