Meta-analysis on the utility of morphometry in the cytological differential diagnosis of thyroid neoplasms

Abstract

Abstract Background: Thyroid nodules, particularly solitary ones, pose a frequent diagnostic challenge. While fine needle aspiration cytology is the primary diagnostic tool for assessing thyroid nodules before surgery, it still exhibits a relatively high rate of false adverse malignancy reports. The Bethesda system categorizes thyroid cytopathology into six categories with specific criteria, yet diagnostic uncertainty often arises, particularly within categories III and IV. Morphometry presents a complementary approach to cytological diagnosis, furnishing valuable diagnostic insights. Nuclear morphometry involves measuring various nuclear parameters through image analysis, offering objective data that may aid lesion classification. Despite its potential, the evaluation of nuclear morphometry in thyroid cytology remains underdeveloped. This study aims to analyze existing literature on the application of nuclear morphometry in differentiating thyroid lesions. Materials and Methods: The literature was searched using the Boolean query “Morphometry and Thyroid Cytology” across Google, PubMed, and Google Scholar. Results: The literature search identified 18 studies that employed morphometry alongside cytomorphology. Among these, three studies were excluded due to being case reports. Image analysis focused on nuclear size parameters such as area, perimeter, maximum and minimum Feret diameters, coefficient of variation of nuclear area, nuclear shape parameters, including shape factor and largest to smallest diameter ratio, and a few nuclear texture parameters. Thirteen studies reported a statistically significant difference in findings between benign and malignant lesions. However, one study indicated that nuclear morphometry might not be a dependable diagnostic tool for assessing thyroid nodules. Conclusion: Computerized nuclear morphometry offers an objective and reproducible means to assess and validate cytological characteristics.