甲状腺结节超声图像数据集
收藏资源简介:
本研究使用了一个包含378个甲状腺结节的超声图像数据集,该数据集由杜克大学收集。数据集中的图像来自不同的超声设备,主要用于评估深度学习算法在甲状腺结节分类中的性能。数据集的创建过程涉及从电子医疗记录中筛选患者,排除不符合条件的病例,最终形成用于分析的图像集。该数据集主要用于医学影像分析,特别是甲状腺结节的良恶性分类,旨在辅助医生进行更准确的诊断。
This study utilized an ultrasound image dataset containing 378 thyroid nodules, which was collected by Duke University. The images in this dataset were acquired from various ultrasound devices and are primarily used to evaluate the performance of deep learning algorithms for thyroid nodule classification. The dataset creation process involved screening patients from electronic medical records, excluding ineligible cases, and finally forming the finalized image set for analysis. This dataset is mainly applied in medical image analysis, particularly for the benign and malignant classification of thyroid nodules, aiming to assist clinicians in making more accurate diagnoses.




