Summer School

Summer School registration is closed.


The 1st ESDIP–MICCAI–ASDP Summer School in Computational Pathology, organized in conjunction with ECDP 2026, offers an intensive two-day training program “From Zero to Hero”. It is intended for computer scientists and AI researchers entering the field of pathology, as well as for pathologists seeking to acquire expertise in computational and machine learning-based methods.
The program combines lectures and hands-on tutorials covering foundational concepts in pathology and machine learning, practical digital image analysis, and advanced topics such as foundation models, model transparency and interpretability, and multimodal data integration.

June 15–16, 2026
Graz Medical University, Austria
Address: Neue Stiftingtalstraße 6, 8010 Graz, Austria.

Please note that registration for the Summer School is separate from registration for ECDP 2026. The fees are as follows:
ESDIP Member - 150 Euros
Non ESDIP Member - 250 Euros
Registration link: https://www.ecdp2026.org/payments/

15th June

Lectured by:
Sertac Kip
Johanna Palacios Ball
Sangjeong Ahn
Yosep Chong
This lecture-based session focuses on the computational researcher and introduces basic terminology and concepts revolving around pathology images. Topics covered will include i) differences between pathology images and real-world images, ii) different types and included artifacts, iii) how they are assessed by a pathologist, and iv) different types of viewers.

Lectured by:
Francesco Ciompi
Anne Martel
This lecture-based session introduces clinicians to the fundamentals of AI with respect to pathology, covering terminology and key concepts from machine learning to deep learning, the foundations of supervised and weakly-supervised learning, as well as interpretability/explainability.

Lectured by:
Tom Bisson
This practical session will introduce the functionality of a widely used open-source whole-slide image (WSI) viewer - QuPath. Attendees will learn how to navigate and customize the software through the use case of cell segmentation in H&E-stained WSIs, followed by basic downstream analysis. Finally, we will briefly transfer these skills to multiplex immunofluorescence images and learn how to export results for use in external tools. [Attendees are expected to have QuPath installed in their own laptop, prior to the session]

Lectured by:
Adam Shephard
Shan Raza
This practical session will demonstrate how Python and TIAToolbox can be used to build complete whole‑slide image analysis pipelines. Participants will learn efficient slide reading, patch extraction, and how to deploy pretrained models for classification and cell detection. [Attendees are expected to have Python and Jupyter Notebooks installed or access to Google Colab for this session.]

Lectured by:
Adam Shephard
Spyridon Bakas

Lectured by:
Adam Shephard
Spyridon Bakas

Norman Zerbe

16th June

Lectured by:
Spyridon Bakas

Rachel Fincham
This lecture-based session will provide a summary of what was presented during the first day of the summer school, getting everyone ready for the second day’s program.

Lectured by:
Anne Martel
Adam Shephard
This lecture-based session introduces WSI‑level classification, starting with patch‑based feature extraction pipelines and moving to multiple‑instance learning (MIL) transformer‑based models. Practical considerations, include benchmarking, implementation, and attention‑based interpretability, are demonstrated through a compact end‑to‑end example, concluding with key limitations and open challenges.

Lectured by:
Jan Obdrzalek
This practical session will demonstrate how attention-based pipelines can generate interpretable visual heatmaps that highlight the image regions driving model decisions in a way that pathologists can meaningfully understand. We will examine XAI techniques such as occlusion or attention-like maps, such as Grad-CAM, their use and limitations. All the techniques will be also applied at the WSI-level. [Attendees are expected to have access to Google Colab for this session. Chromium-based browser (e.g. Google Chrome) is recommended; in particular, Safari is not supported.] [Attendees are expected to have Python and Jupyter Notebooks installed or access to Google Colab for this session.]

This practical session will focus on groups defined during the first day and target model training for distinct computational problems, providing appropriate material in advance. The exact problems to be addressed will be decided after the conclusion of the attendees’ registration and according to their background.

Lectured by:
Sertac Kip
Tom Bisson
Nadieh Khalili
The lecture-based session will focus on introducing technologies beyond traditional histology (such as spatial transcriptomics), as well as beyond pathology alone, towards integrative multi-modal fusion across data types (e.g., radiology, pathology, and molecular characteristics).


Vincenzo L'Imperio
Spyridon Bakas

Sangjeong Ahn
Adam Shephard
Summer school sponsors

Invited Speakers

Adam Shephard

Alessandro Caputo

Francesco Ciompi

Spyridon Bakas

Berna Ozdemir

Johanna Palacios Ball

Anne Martel

Tom Bisson

Petr Holub

Shan Raza

Sanket Kachole

The Summer School in Computational Pathology is an initiative organized by ESDIP, ASDP, and MICCAI, with lead contributions from Norman Zerbe, Sangjeong Ahn, Spyridon Bakas, and Adam Shephard.

This iniciative is further supported by Kip Sertac, Tom Bisson, Vincenzo Della Mea, Mostafa Jahanifar, Johanna Palacios Ball, Petr Holub, Jana Lipkova, Tomas Brazdil, Markus Plass, and Nadieh Khalili.