Fractal dimension analysis of superficial soft tissue lesions on magnetic resonance imaging

We isolated an area of superficial soft tissue lesion margin using Magnetic Resonance Imaging (MRI). In an attempt to distinguish benign tumors cyst and lipoma' from malignant ones'' basal cell carcinoma (BCC) and squamous cell carcinoma (SCC)', objectively and quantatively, we used fractal dimension analysis. The margin of lesion was extracted automatically from the binarized image, and the average fractal dimension was calculated and estimated. As a result, we concluded that the fractal dimension of malignant tumors was greater than that of benign tumors, in spite of no significant difference between BCC and SCC, and cyst and lipoma respectively. Fractal dimension analysis is therefore a useful method of distinguishing benign from malignant tumors when using MRI diagnosis. (author)