Analysis of midbrain area echogenicity in diagnostic transcranial ultrasound images

Abstract

The aim of this paper is evaluating statistical moments potentiality of transcranial sonography (108 subjects) for separation between images obtained from healthy (48 subjects) and Parkinson‘s disease affected (60 subjects) patients. Closer to scanning probe upper half of midbrain with substantia nigra area was selected in three ways as region of interest for analysis: 12, 14, 16 mm diameters of circles. The transcranial sonography images were analysed with 511 combinations obtained from 9 statistical moments. All images were classified with Mahalanobis distance classificator to groups: healthy and PD affected. Efficiency of classification was evaluated with classification rate, sensitivity and specificity.