top of page

Whole-Slide Image Analysis with QuPath

In the tutorial "WSI Analysis with QuPath," we explored the open-source software QuPath, a free platform that provided tools for analyzing scanned histopathological slides, known as whole slide images (WSIs). We learned: (1) How to install QuPath and use it to view and annotate a WSI. (2) How to train the AI-based classification algorithm provided by QuPath so that this algorithm could identify tumor and non-tumor areas on a WSI. (3) How to determine the Ki-67 index using QuPath, a well-known method for assessing cancer cell proliferation in various tumors. No programming or pathology knowledge was required to participate in this tutorial.

Mitosis Detection using Image Processing Techniques in Python

In "Mitosis Detection using Image Processing Techniques in Python," we acquired a fundamental understanding of basic image processing techniques using the example of mitosis detection in PHH3 slides.


We covered the following topics:

- Digital image representation
- Basic operations (color deconvolution, thresholding, morphological      operators, and segmentation)
- Extraction of geometric and statistical features
- Classification of mitosis candidates

Python knowledge was helpful but not essential, as the focus was on image processing concepts.

Histology Toy Example on Machine Learning

In the tutorial "Histology Toy Example Machine Learning," we taught a computer to distinguish between tissue images of two classes. We set up a simple machine learning pipeline that consisted of training, applying, and displaying the results. We learned how to interpret the results, understand the relationship between model complexity and training data, and recognize the importance of testing to identify some pitfalls of binary classifiers. We had to read (and perhaps modify) a few lines of Python code.

bottom of page