Automated Cytopathology of Fine Needle Aspiration for the Detection of Malignancy in Thyroid Cells
Cytopathology of thyroid cells is an established method to detect malignancies in the thyroid. It is of advantage because an anesthesia and a diagnostic laparotomy is not necessary. There are however not yet many studies about automated cytopathology in thyroid cells. To this end, the aim of the present study was to establish an automated diagnosis of malignancy using image analysis and subsequent machine learning and Artificial intelligence. Light microscopy images of 52 patients were analyzed and the results were compared to those of pathology. The results of the automated analysis yielded a sensitivity of 0.94 and a specificity of 0.91 when compared to those of the pathologic diagnoses. The process of machine learning yielded an under curve area of 0.91 as calculated by a ROC-curve. The software used for image analysis, machine learning and classification (diagnosis) are open-source software, respectively.
