Publications by authors named "Haruya Endo"
Article Synopsis
- Variations in histopathology images arise from different staining conditions and imaging devices used in hospitals, which can hinder machine learning models when they encounter unfamiliar data.
- To combat this problem, a new dataset called PathoLogy Images of Scanners and Mobile Phones (PLISM) has been created, including 46 human tissue types and data from 13 different staining and imaging methods.
- The PLISM dataset allows for better assessment of color and texture differences across domains, showing significant variation, especially between whole-slide images and smartphone images, thereby enhancing the reliability of machine learning applications in histological image analysis.
Numerous cancer histopathology specimens have been collected and digitized over the past few decades. A comprehensive evaluation of the distribution of various cells in tumor tissue sections can provide valuable information for understanding cancer. Deep learning is suitable for achieving these goals; however, the collection of extensive, unbiased training data is hindered, thus limiting the production of accurate segmentation models.
View Article and Find Full Text PDF