Paper Overview
- Field: Machine Learning
- Authors: Serena Su, Yifan Wang, Senwei Liang
- Published: 2026-08-17
- arXiv: 2608.16870
- SubSeq strategy: To mitigate data scarcity, the framework randomly extracts informative local trajectory segments during training, encouraging the model to learn from local patterns.
- Interpretability: Gradient-weighted class activation mapping (Grad-CAM) is applied to identify the trajectory features and physical regions of the microfluidic device that drive the model's predictions.
Summary
Accurate classification of circulating tumor cell (CTC) phenotypes can provide valuable information for assessing metastatic potential. Label-free microfluidic devices provide a hydrodynamic obstacle course that transforms subtle biophysical characteristics of CTCs, including size and deformability, into distinct kinematic trajectories. However, the highly nonlinear fluid-structure interactions governing these trajectories make the inverse problem of inferring cellular phenotype from trajectory data analytically intractable.
While deep neural networks (DNNs) have emerged as a powerful approach for addressing this inverse problem, their effectiveness is constrained by the limited availability of trajectory data and the lack of physical interpretability. To address these challenges, the authors propose an interpretable and data-efficient DNN framework for trajectory-based CTC classification.
Key Contributions
Findings
The interpretability analysis shows that local trajectory segments contain substantial biophysical information relevant to accurate classification, highlighting the redundancy of full-length trajectories.
---
*Auto-collected on 2026-08-19*