Thyroid nodule classification in ultrasound imaging using deep transfer learning

Abstract

Our findings demonstrate that the fusion model, which integrates a convolutional neural network (CNN) with traditional machine learning and deep transfer learning techniques, can effectively differentiate between benign and malignant thyroid nodules through the analysis of ultrasound images. This model fusion approach significantly optimizes and enhances diagnostic performance, offering a robust and intelligent tool for the clinical detection of thyroid diseases.

Ähnliche Arbeiten