Abstract

The diagnosis of papillary thyroid cancer is generally based on the findings of intranuclear cytoplasmic inclusions and nuclear grooves. Although anisokaryosis and poikilokaryosis, in papillary thyroid cancer, are not distinct when compared to other cancers, cytological examination can provide useful preoperative information. Our study evaluated the diagnostic role of computer-assisted image analysis for the pre-surgical assessment of papillary thyroid carcinoma. Thyroid aspirates from twenty female patients who were histologically confirmed to have both papillary carcinoma and benign nodules were studied. Different populations of 50 benign cells and 50 malignant cells were analyzed. Five morphometric parameters were selected for analysis: nuclear area, perimeter, maximum length, maximum width and intensity standard variation. The values obtained for papillary carcinomas were higher than the surrounding benign nodules as follows: nuclear area 63.5 vs. 36.1 (p=0.000), nuclear perimeter were 29.4 vs. 22.0 (p=0.000), maximum length 9.6 vs. 7.1 (p=0.000), maximum width 8.2 vs. 6.3 (p=0.000), the ratio between maximal length and maximal width 1.16 vs. 1.13 (p=0.000), the standard variation of intensity 14.9 vs. 15.9 (p=0.101) respectively. Therefore, morphometric information can be helpful for the differential cytological diagnosis of papillary thyroid carcinoma.

• Keywords: Thyroid papillary carcinoma;Aspiration cytology;Image analyzer;Morphometry;