View Proposal

---

Proposer

Marta Vallejo

Title

AI-Based Quantification of TDP-43 Pathology in Fluorescence Microscopy Images of the Spinal Cord (in collaboration with CSIC, Spain)

Goal

The goal of this project is to develop an automated computer vision pipeline to analyse high-resolution fluorescence microscopy images of spinal cord tissue. The system will identify motor neurons and quantify pathological TDP-43 protein accumulation using state-of-the-art image processing and artificial intelligence techniques. The resulting measurements will enable objective comparison of pathological changes between experimental conditions while reducing the need for manual image analysis. The project is conducted in collaboration with the Spanish National Research Council (CSIC), providing access to a unique research dataset. Students producing high-quality results will have the opportunity to contribute to a scientific publication.

Description

Fluorescence microscopy is widely used in neuroscience to investigate the cellular mechanisms underlying neurodegenerative diseases such as amyotrophic lateral sclerosis (ALS). Modern microscopes can generate extremely large multi-channel images containing thousands of cells, making manual analysis slow, subjective and difficult to reproduce. This project aims to develop an automated analysis pipeline capable of extracting quantitative biomarkers from fluorescence microscopy images of spinal cord tissue. The dataset contains multi-channel images including DAPI-labelled nuclei, ChAT-labelled motor neurons and TDP-43 (or phospho-TDP43) protein staining. Possible project tasks include: - preprocessing and enhancement of multi-channel fluorescence images; - tissue and region-of-interest identification; - motor neuron detection and segmentation; - nucleus and cytoplasm segmentation; - automated detection of TDP-43 protein aggregates; - extraction of quantitative biomarkers, including aggregate number, size, intensity and intracellular localisation; - comparison of pathological features between experimental groups; - development of explainable AI visualisations to support biological interpretation Depending on the student's interests and background, the project may focus on classical image processing, deep learning-based segmentation (e.g. U-Net, Cellpose, SAM) or hybrid computer vision approaches. The developed software should provide a reproducible and objective alternative to manual pathology analysis and has strong potential for future translational research. Outstanding projects may lead to a scientific publication in collaboration with researchers at Heriot-Watt University and CSIC.

Resources

Large fluorescence microscopy image dataset provided by the Spanish National Research Council (CSIC).

Background

Neurodegenerative diseases such as amyotrophic lateral sclerosis (ALS) are characterised by abnormal accumulation and mislocalisation of proteins including TDP-43 within motor neurons. Quantifying these pathological changes is essential for understanding disease mechanisms and evaluating potential therapeutic interventions. However, current analyses are often performed manually, making them time-consuming and susceptible to observer variability. Recent advances in computer vision and deep learning have transformed biomedical image analysis, enabling automated segmentation of cells, detection of pathological structures and extraction of quantitative biomarkers. Despite these developments, robust analysis pipelines for large fluorescence microscopy datasets remain an active area of research. This project provides an opportunity to apply modern AI techniques to a challenging real-world biomedical problem while collaborating with an international research team and contributing to research with publication potential.

Url

Difficulty Level

Variable

Ethical Approval

Full

Number Of Students

2

Supervisor

Marta Vallejo

Keywords

computer vision, artificial intelligence, deep learning, biomedical image analysis, fluorescence microscopy, image segmentation, motor neurons, tdp-43, explainable ai

Degrees

Bachelor of Science in Computer Science
Master of Engineering in Software Engineering
Master of Science in Artificial Intelligence
Master of Science in Computing (2 Years)
Master of Science in Data Science
Master of Science in Software Engineering
Bachelor of Science in Computing Science
Postgraduate Diploma in Artificial Intelligence
BSc Data Sciences