
Research Article
Breast Ultrasound Images Clustering Analysis Using Deep Clustering Method
@INPROCEEDINGS{10.1007/978-3-030-94182-6_23, author={Cheng Huang and Jinrong Cui}, title={Breast Ultrasound Images Clustering Analysis Using Deep Clustering Method}, proceedings={IoT and Big Data Technologies for Health Care. Second EAI International Conference, IoTCare 2021, Virtual Event, October 18-19, 2021, Proceedings, Part II}, proceedings_a={IOTCARE PART 2}, year={2022}, month={6}, keywords={Unsupervised learning Clustering analysis Deep clustering Medical data}, doi={10.1007/978-3-030-94182-6_23} }
- Cheng Huang
Jinrong Cui
Year: 2022
Breast Ultrasound Images Clustering Analysis Using Deep Clustering Method
IOTCARE PART 2
Springer
DOI: 10.1007/978-3-030-94182-6_23
Abstract
With the expanding volume of medical data, the supervised algorithms have the difficulty in obtaining labels of the medical data. To address this problem, many researchers have introduced unsupervised learning algorithms to analyze medical data and to mine the potential information among different samples. Simultaneously, with the brisk growing of deep learning, the association of unsupervised algorithms and deep learning is becoming further intimate. In this paper, we will use clustering algorithm which is commonly used in unsupervised learning to analyze breast ultrasound images. The Improved Deep Embedded Clustering (IDEC) algorithm will be introduced in this paper, which is broad used in the current deep clustering methods to do clustering analysis. Meanwhile we have added the graph embedding strategies in IDEC algorithm to enhance the clustering performance. The algorithm includes two phases: pre-training phase to capture hidden features and fine- tuning phase to enhance clustering result. We have exerted our algorithm on breast ultrasound images dataset to perform clustering analysis, and the experiment results indicate that the clustering performance of this algorithm is better than other traditional clustering algorithms.