Computerized Cytological Features for Papillary Thyroid Cancer Diagnosis-Preliminary Report

Area of Science:

• Cytopathology
• Oncology
• Medical Imaging

Background:

• Fine needle aspiration cytology (FNAC) is crucial for diagnosing thyroid nodules before surgery.
• Indeterminate FNAC results pose a diagnostic challenge, necessitating improved accuracy.
• Computerized analysis of cytological features offers potential for enhanced diagnostic precision.

Purpose of the Study:

• To investigate the utility of quantified cytological features in improving the diagnosis of indeterminate thyroid nodules.
• To develop a computer-aided tool for differentiating benign goiters from papillary thyroid cancer (PTC).

Main Methods:

• Retrospective cross-sectional study of 240 thyroid nodule cases with known histologic diagnoses.
• Quantification and analysis of morphological and chromatic features from FNAC samples.
• Multivariate analysis to identify significant cytological features differentiating PTC from benign goiters.

Main Results:

• Six quantified cytological features were significantly different between PTC and benign goiter groups.
• The developed model achieved an Area Under the Receiver Operating Characteristics (AUROC) of 81.3% for atypia, 78.7% for suspicious for malignancy, and 56.8% for follicular neoplasm.
• The computer-aided tool demonstrated potential in estimating malignancy risk for indeterminate FNAC results.

Conclusions:

• Quantification of cytological features is a promising method for enhancing thyroid nodule diagnosis.
• A computer-aided tool based on these features can aid in diagnosing PTC in indeterminate FNAC cases.
• This approach may lead to more accurate preoperative diagnosis and better patient management.