PIDGIN Project: Harnessing AI and HPC to Advance Lung Cancer Diagnosis

Lung cancer remains one of the most common and deadly cancers worldwide. Early and accurate diagnosis is critical to improving treatment outcomes. The PIDGIN Project at Hasselt University is taking an innovative approach to this challenge—combining pathology, biomedical sciences, oncology, and data science to develop a more precise diagnostic method based on tumor heterogeneity.

This interdisciplinary collaboration brings together researchers, clinicians, and data scientists, leveraging cutting-edge artificial intelligence (AI) and high-performance computing (HPC) to transform how tumor samples are analyzed.

From Microscopes to Digital Slides

Traditionally, pathologists examine microscopic tissue slides to assess tumors and determine their type. While effective, this method relies on human observation, which can sometimes miss subtle differences between healthy and cancerous cells.

The PIDGIN Project uses advanced slide scanners to digitize high-resolution images of these tissue samples. These digital slides offer several key advantages:

• Remote accessibility: Researchers can access and share slides from anywhere.

• Computational analysis: Images can be processed using AI algorithms to detect patterns invisible to the human eye.

• Objective and reproducible measurements: Digitization allows for automated, standardized analysis across multiple samples.

AI at the Cellular Level

One of the core AI techniques used in the PIDGIN Project is the U-Net algorithm, a type of convolutional neural network designed for image segmentation. It detects individual cells within the high-resolution tumor images, mapping their exact location and structure.

But identifying cells is only the first step. The project also uses sophisticated tissue colorings that reveal the “ground truth” of cell types—for example, whether a cell is a blood vessel or another structure. This data is then used to train artificial neural networks to recognize cell types directly from the high-resolution images, eliminating the need for repeated, costly and time-consuming staining.

Quantifying Tumor Heterogeneity

By extracting the location, features, and neighborhood of every cell, the team develops statistical models to quantify tumor heterogeneity—the variation in cell types and their spatial distribution.

Understanding this heterogeneity is crucial. It can reveal patterns linked to disease progression, predict how tumors may respond to treatment, and guide personalized therapeutic strategies.

The Role of VSC Infrastructure

The project’s success depends heavily on high-performance computing and large-scale data storage. The Flemish Supercomputing Centre (VSC) has been a key enabler by providing:

1. Tier-1 Data Storage – Secure and scalable infrastructure to store the high-volume digital slide images.

2. Tier-1 Compute Power – Resources to train deep neural networks capable of processing complex image datasets.

3. Large-Memory Nodes – Essential for handling the size and complexity of high-resolution tumor images.

“VSC gave us a great solution to store and share our data, but also to perform complex computations at the same time,” the team notes.

Meet the Researchers

The PIDGIN Project is driven by an interdisciplinary team at Hasselt University:

• Esther Wolfs, Associate Professor

• Dirk Valkenborg, Professor

• Christel Faes, Professor

Watch the Success Stories

🎥 Current story | AI in Lung Cancer Diagnosis: here

🎥 First story | Data Entry & Data Retrieval: here

---

🔍 Your Research Matters — Let’s Share It!

Have you used VSC’s computing power in your research? Did our infrastructure support your simulations, data analysis, or workflow?

We’d love to hear about it!

Take part in our #ShareYourSuccess campaign and show how VSC helped move your research forward. Whether it’s a publication, a project highlight, or a visual from your work, your story can inspire others.

🖥️ Be featured on our website and social media. Show the impact of your work. Help grow our research community

📬 Submit your story: https://www.vscentrum.be/sys