Examination of digital medical images, either radiological or ultrasound, is based mainly on comparison between normal and pathologic appearances. This comparison is performed usually by a human examiner with a long experience gained over a considerable amount of time. To minimize the degree of subjectivity of the examiner it is necessary to use independent user methods, with high degree of objectivity, such as mathematical and statistical methods.
In this study the goal was to quantify mathematically a total number of 13 radiography and 24 ultrasound images using ImageJ 1.46q software (Wayne Rasband, National Institutes of Health, Bethesda). The two categories of images contained both normal and pathological aspects of the analyzed structures. The images were investigated regarding their fractal dimension, lacunarity and histogram.
The data recorded from measurements were statistically analyzed, and final conclusions were drawn on the similarities and differences between structure samples.
fractal dimension, lacunarity, histogram, ImageJ
Andrej Stefan's research interest belongs to applications of computer/mathematical analysis technics in diagnostics (histopathology, ultrasonography, x-ray etc.), in order to offer objective measurements of structures and patterns.
Presenting author: Andrei Stefan
Organisation: Department of Pathophysiology, Faculty of Veterinary Medicine, University of Agricultural Science and Veterinary Medicine, Cluj-Napoca, Romania