Thyroid Cancer Diagnosis from Images based on Deep-Transfer Learning

The thyroid gland is an important organ of the human body that controlsthe body's metabolic operations. Improper thyroid hormone productioncan cause many complications. Early detection of thyroid disorders iscritical to prevent such complications. Previous work on the diagnosis ofthyroid disease, the problem of class imbalance has not been wellinvestigated. In addition, the existing studies mainly focus on the binaryclass problem. The aim of this study is to solve these problems with theproposed approach. This study uses an optimized deep-transfer learningalgorithm to find optimal parameters to obtain higher performance forthyroid disease diagnosis. The experimental results show that theaccuracy of ۹۸.۷۴% can be obtained using the optimized deep-transfermodel.

Keywords:

Thyroid Cancer , Deep-Transfer Leaning , FCNN