Related Experiment Video
Updated: Nov 25, 2025
09:43
Author Spotlight: Unveiling the Polyfunctionality and Heterogeneity in Immune Responses
Published on: March 8, 2024
2.1K
Improved efficiency of urine cell image segmentation using droplet microfluidics technology
Shuxing Lv1, Yuying Chu1, Panpan Zhang2
1School of Medical Laboratory, Tianjin Medical University, Tianjin, China.
Summary
Encapsulating urine cells in droplets using microfluidics significantly improves machine vision segmentation accuracy. This method reduces interference from overlapping cells and other elements, enhancing biological sample recognition.
Area of Science:
- Biomedical Engineering
- Computer Vision
- Microfluidics
Background:
- Machine vision is crucial for biological sample recognition, with image segmentation being a key step.
- Interference from overlapping cells and other urine elements hinders accurate machine vision segmentation.
Purpose of the Study:
- To improve machine vision segmentation accuracy for urine cell images.
- To reduce interference factors like cell overlap and crystallization using microfluidics.
Main Methods:
- A urine cell droplet microfluidic chip system was used to encapsulate single cells.
- Analysis of cell overlap, salt crystallization interference, and segmentation performance using the Otsu algorithm.
- Evaluation of segmentation accuracy using Dice, Jaccard, precision, and recall metrics.
Main Results:
- Droplet encapsulation effectively isolated urine cells from interfering elements.
- The proposed microfluidic method improved image segmentation without algorithm optimization.
- Quantitative metrics showed enhanced segmentation performance for encapsulated cells.
Conclusions:
- Microfluidic droplet encapsulation is an effective strategy to enhance urine cell image segmentation.
- This approach offers a pathway to improve machine vision accuracy in biological sample analysis.
- The findings have implications for automated cell detection and research applications.
