Detection of Aorta Anatomical Structures Characterized by Various Levels of Pixel Intensity

E. Sobotnicka, J. Wróbel, A. Sobotnicki (Inst. of Medical Techn. and Equipment, Poland)

The paper discusses image processing which takes into account the split/merge method. Owing to the automation of the method, it was possible to split elongated anatomical structures (the aorta in the case under scrutiny) into sections which consist of regions containing pixels of the same intensity level. The analyzed image was initially processed then filtered, respectively. The segmentation stage, which took place afterwards, was followed by a statistical analysis and interpretation of image features. The paper presents a method of obtaining significant numerical and symbolic information in image processing.