Evaluation parameters and technique for classification of intraocular tumours

Abstract

Melanoma, metastases and no malignant tumour are frequently observed types of intraocular tumours. For each third patient the direct reason of death causes eye melanoma. Therefore early diagnostic of this pathology is crucial for proper treatment and prognosis. Ultrasound echographic technique for visualization intraocular tumours is widely used. The possibility to apply ultrasound technique for classification of various tumours is based on the direct relationship between tumour tissue density distribution and its morphology. Diagnostic examinations with ultrasonic diagnostic imaging system Mentor Advent, using A/B scan mode are performed in the laboratory of Ophthalmology of Kaunas Medicine University Hospital. Diagnostic images of this examination are transmitted from video output of ultrasound system to PC by the use of frame grabber. Captured echographic images were analyzed using numerical computation and visualization software. Classification of intraocular tumours can be performed using these images. Classification parameters obtained from diagnostic images are: tumours height, base, shape, choroidal excavation, regularity and reflectivity. Influence of these parameters to classification reliability was one of the goals of this investigation. The digitized values of reflectivity coefficient (low, medium, high) and irregularity coefficient (regular, irregular) were used. The optimal values of thresholds for the reflectivity coefficient and irregularity coefficient were found.