Scholarlink

This paper explores the application of machine learning techniques to classify circulating tumor cells (CTCs) based on their trajectories within a hyperuniform micropost microfluidic device. Leveraging cell-based modeling, we generated a synthetic dataset simulating the dynamics of CTCs in blood flow. Three machine learning architectures were applied to analyze the trajectory data: a Convolutional Neural Network (CNN), a hybrid model combining CNNs with long short-term memory (LSTM) networks, and the eXtreme Gradient Boosting (XGBoost) algorithm. These models achieved an average classification accuracy of 80% in distinguishing between different CTC phenotypes, highlighting the potential of this approach for early cancer detection. All code implementations are available as open source at: https://github.com/imsanjoykb/Microfluidic-Device-Data-CTC_Model.git .

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Classification of Circulating Tumor Cells Using Machine Learning on Microfluidic Trajectory Data

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Sanjoy Kumar,
  • Yifan Wang,
  • Huixin Zhan,
  • Karl Gardner,
  • Travis Thompson,
  • Wei Li,
  • Suncica Canic

摘要

This paper explores the application of machine learning techniques to classify circulating tumor cells (CTCs) based on their trajectories within a hyperuniform micropost microfluidic device. Leveraging cell-based modeling, we generated a synthetic dataset simulating the dynamics of CTCs in blood flow. Three machine learning architectures were applied to analyze the trajectory data: a Convolutional Neural Network (CNN), a hybrid model combining CNNs with long short-term memory (LSTM) networks, and the eXtreme Gradient Boosting (XGBoost) algorithm. These models achieved an average classification accuracy of 80% in distinguishing between different CTC phenotypes, highlighting the potential of this approach for early cancer detection. All code implementations are available as open source at: https://github.com/imsanjoykb/Microfluidic-Device-Data-CTC_Model.git .