Haralick Texture Analysis for Differentiating Suspicious Prostate Lesions from Normal Tissue in Low-Field MRI.

Area of Science:

• Radiology and Medical Imaging
• Biomedical Engineering
• Oncology

Background:

• High-field MRI for prostate cancer detection is effective but costly.
• Low-field MRI offers a more accessible and cost-effective alternative.
• Haralick texture analysis is a quantitative method to analyze image texture.

Purpose of the Study:

• To evaluate the feasibility of Haralick texture analysis on low-field, T2-weighted MRI for prostate cancer detection.
• To extend the application of texture analysis from high-field to low-field MRI systems.
• To assess the potential of Haralick features in differentiating cancerous from non-suspicious prostate tissue.

Main Methods:

• Twenty-one patients with biopsy-proven prostate cancer underwent low-field MRI.
• Suspicious regions of interest (ROIs) from high-field MRI were registered to low-field images.
• Four Haralick texture features (Energy, Correlation, Contrast, Homogeneity) were extracted from cancerous and normal ROIs using two methods.

Main Results:

• Statistically significant differences were found in Haralick texture features between cancerous and non-suspicious prostate regions.
• Energy and Homogeneity were significantly elevated in cancerous ROIs (p < 0.00001–0.004).
• Contrast and Correlation were significantly reduced in cancerous ROIs (p < 0.00001–0.03).

Conclusions:

• Haralick texture analysis is feasible and informative for prostate cancer detection using low-field MRI.
• Texture features can effectively differentiate cancerous from normal prostate tissue on low-field MRI.
• This approach holds promise for improving accessible prostate cancer diagnosis.