Texture Feature Differences between Fresh and Frozen-Thawed Ex-Vivo Porcine Liver Tissue in B-mode Ultrasonic Imaging

BACKGROUND: Ultrasound guidance of cryotherapy could be improved if changes of tissue characterization parameters during freezing could be used to monitor freezing profiles and estimate frostbite distributions. OBJECTIVE: To investigate if B-mode ultrasonic imaging can provide sufficient information to define the damaged zone during cryotherapy by quantifying the differences of texture features between fresh and frozen-thawed porcine liver tissue. MATERIALS AND METHODS: The B-mode ultrasonic images were obtained from ten porcine liver tissue samples before and after frozen-thawing treatment, then 20 texture parameters of the grey level histogram (GLH) and the grey level co-occurrence matrix (GLCM) from each image were extracted and analyzed statistically. RESULTS: It was found that the differences in two texture parameters between the fresh and frozen-thawed tissue are significant enough for diagnostic purpose (variance: 317.0%, contrast: 144.4%). CONCLUSION: Texture feature analysis of B-mode ultrasonic images can effectively differentiate the fresh and frozen-thawed porcine liver tissues, which offers a prospect for the future practice of evaluating the effectiveness of cryotherapy using Ultrasound.