Quantitative classification of breast fine needle aspirates using the AxioHOME system

Abstract

Fine needle aspiration cytology is useful in the pre-operative assessment of patients with breast lumps. Lesions are reported as benign, suspicious or malignant. The number of suspicious categories is high in inexperienced hands thus limiting this useful diagnostic tool. The aim was to evaluate quantitative methods of classifying breast fine needle aspirates using the Highly Optimized Microscope Environment system. May- Grumwald Giemsa-stained archived slides were retrieved and smear quality assessed. Fifty epithelial cells corresponding to the cytological grading on each slide were measured using the system\'s general morphometry program. Generated data was exported to Microsoft Excel for analysis. A significant difference in the mean nuclear area, mean nuclear perimeter, mean largest nuclear diameter was found between the slides graded as benign, suspicious malignant and malignant. C2 & C4, D 1, P < 0.004. C4 & C5, Area P < 0.001. The means of
the number count of nucleoli was able to distinguish between FNAC Benign and Malignant category C2/C5, P = 00440 and suspicious benign and malignant C3/C5 P = 0.00486. Morphometry could be useful in situations where experienced cytopathologists are unavailable especially when this program can also measure the degree of dispersion between cells

Keywords: Morphometric image analysis, aspiration cytology, breast.

International Journal of Biological and Chemical Sciences Vol. 2 (2) 2008: pp. 139-145