Matthew Lyon

Matthew Lyon

PhD Student

University of Manchester

I am currently a PhD student in the Machine Learning group at the University of Manchester. My research focuses on improving MRI data through deep learning, and incorporating geometric priors into deep learning models.

Previously, I received my masters of Medical Physics from the University of Sydney, and a Bachelors of Science (Honours) in Physics at the University of Warwick. I have experience working as a research software engineer at several research institutes.

Download my CV.

Interests
  • Deep Learning
  • Computer Vision
  • Geometric Deep Learning
  • Diffusion MRI
Education
  • PhD in Computer Science, 2024

    University of Manchester

  • Masters in Medical Physics, 2016

    University of Sydney

  • BSc (Hons) in Physics, 2014

    University of Warwick

Experience

 
 
 
 
 
University of Manchester
Research Assistant
Jun 2022 – Present Manchester
  • Designed and implemented data cleaning and preprocessing pipelines.
  • Performed exploratory analysis on large time-series datasets.
  • Lead tutorials on several machine learning courses.
  • Graded assignments and exams for several machine learning courses.
 
 
 
 
 
Save Sight Institute
Research Software Engineer
Aug 2019 – Aug 2020 Sydney
  • Developed, tested, and documented neuroimaging processing pipelines.
  • Lead algorithm design and optimization workflows.
  • Consulted on neuroimaging analysis techniques and signal processing.
 
 
 
 
 
Sydney Neuroimaging Analysis Centre
Neuroimaging Analyst
Aug 2019 – Jan 2020 Sydney
  • Developed, implemented, and led QC on neuroimaging analysis pipelines.
  • Conducted exploratory data analyses.
 
 
 
 
 
Heart Research Institute
Research Software Engineer
Jul 2017 – Jul 2019 Sydney
  • Built and managed a distributed computing cluster.
  • Developed, tested, and documented neuroimaging processing pipelines.
  • Oversaw data ingestion and QC/QA, created dashboard visualisations.
  • Conducted clinical research using MRI data.

Publications

(2023). Spatio-Angular Convolutions for Super-resolution in Diffusion MRI. In NeurIPS 2023.

Cite DOI

(2022). Angular Super-Resolution in Diffusion MRI with a 3D Recurrent Convolutional Autoencoder. In MIDL 2022.

PDF Cite DOI

(2019). Structural core of the executive control network: A high angular resolution diffusion MRI study. In Human Brain Mapping.

PDF Cite DOI

(2019). Profound and reproducible patterns of reduced regional gray matter characterize major depressive disorder. In Translational Psychiatry.

PDF Cite DOI

(2019). Is occipital bending a structural biomarker of risk for depression and sensitivity to treatment?. In Journal of Clinical Neuroscience.

PDF Cite DOI

(2019). Gender-specific structural abnormalities in major depressive disorder revealed by fixel-based analysis. In NeuroImage: Clinical.

PDF Cite DOI

Machine Learning

Technologies

TensorFlow, Keras, PyTorch, NumPy, pandas

Tasks

Classification, Segmentation, Synthesis, Probabilistic Modelling, Unsupervised Learning

Models

CNN, ResNet, UNet, ViT, GAN, Transformer, RNN, Autoencoder