CellaVision - RAL Diagnostics logo

Master’s Thesis: Finding the Best Areas for Automated Blood Cell Analysis

CellaVision - RAL Diagnostics

Lund, SwedenOnsitePosted 28 days ago
InternshipInternBioinformatics & Data ScienceDiagnostics
Original posting ↗

Summary

How can an automated microscope know which part of a blood smear is best suited for analysis? At CellaVision, we develop digital microscopy solutions that support laboratories in analyzing blood cells. Choosing the right region of a slide is an important part of producing reliable results.

Job description

Master's Thesis Project: Finding the Best Areas for Automated Blood Cell Analysis

How can an automated microscope know which part of a blood smear is best suited for analysis? At CellaVision, we develop digital microscopy solutions that support laboratories in analyzing blood cells. Choosing the right region of a slide is an important part of producing reliable results.

Morphologists consider several characteristics when deciding which regions of a blood smear are suitable for analysis. These may include how easily individual red blood cells can be distinguished, their distribution across the image, the amount of empty background, and visible cell features. We want to investigate how these observations can be translated into an automated, transparent and reliable method for assessing different areas of a blood smear.

The Assignment

In this thesis, you will explore how image characteristics such as overlapping red blood cells, central pallor, cell distribution, and background regions can help identify suitable areas for automated analysis. You will first investigate the problem using high-resolution images, where more detailed information is available, and then explore how to obtain a comparable assessment from low-resolution images to support efficient scanning.

What You Will Do

  • Investigate how suitable analysis regions can be characterized and assessed, including a review of relevant literature and/or discussions with experienced morphologists and other domain experts.
  • Identify and evaluate image characteristics that distinguish suitable and less suitable blood smear regions.
  • Develop and evaluate an explainable image-analysis method, with machine learning as a central component, based on blood smear images.
  • Explore how insights from high-resolution images can be translated into efficient assessments using low-resolution images.
  • Evaluate the results against existing approaches and, where possible, assessments by experienced morphologists.

Who Are You

One or two students pursuing a Master's degree in Computer Science, Engineering Mathematics, Biomedical Engineering, or a related field. You enjoy analytical problem-solving and programming and have experience with machine learning. An interest in image processing and computer vision is also valuable.

What You Can Expect

You will work on a real medical imaging challenge together with experienced engineers at CellaVision, combining algorithm development with practical evaluation. The goal is to produce a prototype and insights that may contribute to future improvements in automated blood cell analysis.

Thesis level: Master’s thesis (30 ECTS). The project will be carried out at our CellaVision HQ in Lund, Sweden—a modern office with great infrastructure and commuting options.

Requirements

  • Pursuing a Master's degree in Computer Science, Engineering Mathematics, Biomedical Engineering, or a related field
  • Enjoys analytical problem-solving
  • Programming experience
  • Experience with machine learning
  • Interest in image processing and computer vision is valuable

Company overview

CellaVision - RAL Diagnostics is a French manufacturer of biological stains and reagents for sample preparation in hematology, pathology, cytology, bacteriology, and parasitology. Based in Martillac, near Bordeaux, the company produces staining reagents and instruments—including the RAL Smearbox, Stainbox, and Stainer systems—that enable standardized, high-quality preparation of blood and tissue samples for microscopic and digital analysis. Its reagents help identify cell and tissue morphology, parasites, and bacteria essential for diagnosing a wide range of diseases, and are supplied as liquids or ready-to-use kits. Beyond clinical diagnostics, RAL Diagnostics also serves pharmaceutical, cosmetic, and industrial customers seeking raw materials such as eosin, methylene blue, and fluorescein. Founded in 1916, the company was acquired by Swedish medical technology firm CellaVision AB in October 2019, integrating its sample-preparation portfolio with CellaVision's AI-driven digital microscopy analyzers. RAL Diagnostics employs roughly 70 people and operates a dedicated reagent production facility in France, supporting CellaVision's global hematology solutions across more than 40 countries.

Explore more jobs of your matchFind more biopharma roles with GetBioJob.