Studies on tissue characterization by texture analysis with co-occurrence matrix method using ultrasonography and CT...

Area of Science:

• Medical Imaging
• Image Analysis
• Radiology

Background:

• Texture analysis is a quantitative method for image assessment.
• Co-occurrence matrix is a common technique for texture analysis.

Purpose of the Study:

• To evaluate the efficacy of texture analysis using the co-occurrence matrix method.
• To analyze ultrasonograms of normal and diseased livers.
• To analyze X-ray CT images of normal lungs and lungs with idiopathic interstitial pneumonia.

Main Methods:

• Texture analysis with co-occurrence matrix method was applied to ultrasonograms and CT images.
• Subjects included normal, fatty, and cirrhotic livers, and normal lungs and lungs with idiopathic interstitial pneumonia.
• Image analysis was performed under standardized conditions.

Main Results:

• While normal and diseased liver groups showed no significant difference, variations in fibrosis grade and nodule size in cirrhotic livers yielded different Fisher ratios.
• Significant differences were observed between normal lungs and lungs affected by idiopathic interstitial pneumonia.
• Texture analysis demonstrated potential in differentiating pathological grades in liver lesions and classifying lung CT images.

Conclusions:

• Texture analysis using the co-occurrence matrix method can differentiate between various pathological grades in liver and lung diseases.
• This technique shows potential for classifying medical images based on pathological characteristics.
• Further research may establish texture analysis as a valuable tool in diagnostic imaging.