Skip to content
View leandervaneekelen's full-sized avatar

Highlights

  • Pro

Block or report leandervaneekelen

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
leandervaneekelen/README.md

Hi, I'm Leander 👋

AI Scientist · Medical Image Analysis · Research Software Engineer

LinkedIn Email

I enjoy building and deploying deep learning models for (bio)medical applications — from research prototype to clinical practice. My PhD research focused on predicting immunotherapy outcomes in lung cancer patients using computational pathology. After my PhD, I was a research software engineer implementing AI in the clinical practice of the Radboud University Medical Center in the Netherlands.

From March 2026, I am looking for a new job as an AI scientist/engineer or research software engineer.

🔬 Research Highlights

PD-L1 Cell Detection in Non-Small Cell Lung Cancer

Paper Code Demo

Developed a deep learning pipeline to quantify PD-L1 expression at the cell level in whole-slide histopathology images, a key biomarker for immunotherapy eligibility.

PD-L1 detection result


Open Histopathology Dataset for Detection & Segmentation

Paper Code Data

Co-created a publicly available H&E and PD-L1 annotated dataset for training and benchmarking AI models in computational pathology.

PD-L1 detection result


Cellularity Quantification in Bone Marrow Biopsies

Paper
Built a segmentation model for automated quantification of cellularity in bone marrow histology to measure how cellularity varies throughout life.

PD-L1 detection result

🛠️ Building

Nexum: a scalable ML serving platform for Radboudumc

As a research software engineer (RSE) at Radboudumc from February 2025 to March 2026, I helped build Nexum, a containerised framework for connecting our PACS/IMS to a compute backend. Nexum was designed to enable inference of any containerised algorithm on anonymized slides in a scalable and anonymous manner. To achieve this, we built a Docker Compose stack consisting of a broker and compute workers coupled to an asynchronous Celery queue. The workers could be horizontally scaled by launching more GPU instances on whatever (cloud) hardware is available. Nginx served as the central reverse proxy, routing traffic from external clients (the QA officer-facing dashboard and the PACS) to the internal services. Job metadata and results were persisted in MongoDB, accessible via a Mongo Express admin interface.

Diagram of Nexum's containers Screenshot of algorithm visualization in PACS viewer

🎯 Skills

AI & Machine Learning
PyTorch scikit-learn OpenCV W&B

Infrastructure & Deployment
Python Docker FastAPI MongoDB Redis Celery

Practices
Git CI/CD HPC


🏢 Experience

Period Role Organisation
2025 – 2026 Research Software Engineer Radboudumc, Dept. of Pathology
2021 – 2025 PhD Researcher, Medical Image Analysis Radboudumc, Computational Pathology
2020 – 2021 AI Research Intern Polytechnique Montréal

Popular repositories Loading

  1. forked-pdb forked-pdb Public

    Forked from Lightning-AI/forked-pdb

    Python pdb for multiple processes

    Python 1

  2. essential-skills essential-skills Public

    Forked from tueimage/essential-skills

    Educational module 'Essential Skills for Machine Learning'

    Python 1

  3. prov-gigapath prov-gigapath Public

    Forked from prov-gigapath/prov-gigapath

    Prov-GigaPath: A whole-slide foundation model for digital pathology from real-world data

    Python 1

  4. 8DM20-CS-Prostate-Group-4 8DM20-CS-Prostate-Group-4 Public

    Repository for project work of Group 4 on 8DM20, Capita Selecta in Medical Image Analysis

    Python

  5. Deedy-Resume Deedy-Resume Public

    Forked from anthonyattard/Deedy-Resume

    A one page, two asymmetric column resume and cover letter template in XeTeX that caters to an undergraduate Computer Science student

    TeX

  6. pydensecrf pydensecrf Public

    Forked from lucasb-eyer/pydensecrf

    Python wrapper to Philipp Krähenbühl's dense (fully connected) CRFs with gaussian edge potentials.

    C++