Master’s Thesis (6 months) – Machine Learning for Cytology Image Analysis

12.03.2026, Abschlussarbeiten, Bachelor- und Masterarbeiten

Bronchoalveolar lavage cytology is widely used to assess lung diseases, infections, and cancer-related changes, but the analysis is typically performed manually by experts. We are seeking a Master’s student for a thesis project focused on machine learning and computer vision methods for bronchoalveolar lavage (BAL) cytology analysis to automate this analysis.

Project

Usually, human and veterinarian lung lavage cytology is manually assessed by so-called bronchoalveolar lavage (BAL) cell differentiation at the microscope, which is time-consuming and highly variable and examiner dependent.
The goal of this master thesis is to establish a machine-learning based lung BAL cytology tool for quantitative cell differentiation of murine samples.

The work will include:
● Developing neural cellular automata models for single-cell detection, segmentation and classification in scanned BAL cytology images
● Evaluating existing cytology expert models and foundation models, including in-house developments
● Implementing active learning strategies to guide efficient manual annotation
● Benchmarking model performance for BAL cytology cell type recognition

Tools and Methods
● PyTorch for machine learning model development
● Fiji (ImageJ) for image processing and annotation
● Active learning pipelines for annotation-efficient model training
● Computer vision methods for microscopy data

Requirements
● Master’s student in bioinformatics, computer science, physics, applied math, biomedical engineering, or a related field
● Experience with Python
● Interest in machine learning and image-based biological data
● Experience with deep learning frameworks (PyTorch, TensorFlow) is a plus

Practical Information
● Duration: ~6 months
● Start date: Flexible
● Location: Oberschleissheim, Helmholtz Munich, Institute of Computational Biology

To apply, please send a CV and a short statement of interest to nikita.moshkov@helmholtz-munich.de.

Kontakt: nikita.moshkov@helmholtz-munich.de