Research

Research Interests

My research interests focus on two scientific areas:

  1. Understanding how the retina learns to processing visual stimuli.
  2. The interface of hardware, software and wetware in the pursuit of high quality, quantitative bioimaging.

Both of these areas requires the integration of skills from computing, engineering, chemistry and the life sciences. As such, my research career has moved me between disciplines and involves a large number of cross-disciplinary collaborations.

You can find my publications here.

Current Research

How can babies see from the moment their born? How does the retina learn to process images before sending them to the brain? My current research is aimed at understanding these fundamental neuroscience questions by combining state-of-the-art light sheet microscopy, image analysis and network science.

I am doing this in the zebrafish (Danio Rerio), a popular model organism in microscopy due to it’s optical transparency and the wide range of fluorescence transgenic lines now available. Zebrafish are even more attractive as models for retinal developmental biology due to their relative simplicity but core similarities to mammalian retinal development.

Past Projects

There are many interesting challenges in quantitative bioimaging including the reductions of photodamage, low signal-to-noise and images with occluded or incomplete information. These complications often have to be solved in a scenario-specific manner using a combination of bright optics (pun intended) and computational analysis. My Masters project, PhD research and research assistant work all focussed on these custom solutions.

Hybrid Optical Gating in Light Sheet Microscopy

As Research Assistant to Dr Jonathan Taylor (Glasgow University) I worked on real-time image processing for the synchronisation of 3D images of the in vivo, beating zebrafish heart. Through this project we have developed a technique we call hybrid optical gating.

The constant motion of the heart is a major obstacle to live imaging and, in the past, fish have been fixed or their heartbeats have been pharmaceutically slowed down. Our research aims to develop the hardware and software needed to create truly synchronised 3D images of whole heartbeats over long time periods so that the long term effects of drugs or injury on heart and vasculature can be elucidated.

This research was funded by the British Heart Foundation.

A preprint of this work can be found on bioRxiv.

Vessel Enhancement

Bowler Hat Transform enhancement of vessels in retinal images.
The Bowler Hat Transform enhances vessels in retinal images – this is an important first step in computer-aided diagnosis.

Working with Cigdem Sazak and Boguslaw Obara (Durham University) I developed a new approach for the enhancement of vessel-like objects in biological and medical imaging called the bowler-hat transform. We have used this approach in the analysis of fungal and slime mould networks, with our collaborator Mark Fricker (Oxford University), retinal imaging and brain vasculature.

My contribution to this work was funded by the EPSRC.

The key output of this work was published in Pattern Recognition.

Automated Nuclei Detection

Research - Nuclei detected by HEDAR

At Durham University, my PhD research focussed on the use of mathematical morphology to quantify biological and medical images. One such challenge was the automated detection of ellipse-like objects, such as cell nulcei, in fluorescence microscopy. Working with Philip Jackson and Boguslaw Obara, I developed a new approach to identifying ellipses based on Hilbert-edge detection and ranging (HEDAR) and assumptions based on the shape on an ellipse.

My contribution to this work was funded by EPSRC.

A publication is currently under preparation.

Object-based Colocalisation

Research - Results from object-based colocalisation in plant cells

Working with researchers at Durham University, the University of Essex and the University of Exeter, I published an article focussing on the benefits of using object-based approaches to quantify colocalisation experiments. The article highlighted not only the benefits of such an approach but the importance of understanding the underlying algorithms and their assumptions by showing the biases and uncertainties of the approach.

My contribution to this work was funded by EPSRC.

The key output of this work was published in Functional Plant Biology.