Related Experiment Video
Updated: Dec 6, 2025
06:44
Automated Dissection Protocol for Tumor Enrichment in Low Tumor Content Tissues
Published on: March 29, 2021
2.8K
TissueWand, a Rapid Histopathology Annotation Tool
Martin Lindvall1,2,3, Alexander Sanner1, Fredrik Petré1
1Sectra AB, Research Department, Linköping, Sweden.
Journal of Pathology Informatics
|October 12, 2020
Summary
TissueWand is a new tool that speeds up the annotation of histopathology images. This interactive software assists experts in creating training data for machine learning models, improving efficiency without compromising quality.
Area of Science:
- Histopathology
- Machine Learning
- Computational Pathology
Background:
- Machine learning (ML) advancements offer potential for diagnostic tools in histopathology.
- Current ML approaches require extensive expert-annotated training data, a time-consuming process.
- Efficient tools are needed to aid annotation, saving time and maintaining data quality.
Purpose of the Study:
- To develop an interactive tool for efficient annotation of large histopathological images.
- To address the need for faster dataset curation in the absence of task-specific ML models.
Main Methods:
- Iterative design process to develop the TissueWand tool.
- Focus on interactive concepts and semi-automation with rapid local feedback.
- Evaluation through a user study comparing to manual annotation.
Main Results:
- TissueWand enables efficient annotation of gigapixel histopathological images.
- Semi-automation based on rapid interaction feedback is a key design component.
- User study demonstrated substantial speed-up and maintained quality compared to manual methods.
Conclusions:
- TissueWand shows potential to enhance early-stage dataset curation for histopathology.
- The tool can replace manual annotation methods when no specific ML model is yet available.
- Facilitates the development of ML-assisted diagnostic tools in pathology.
