Skip to content
Soheil Fallah

Soheil Fallah

AI and machine learning for medical imaging

I work on artificial intelligence and machine learning for medical imaging and healthcare. I began in IT, spending five years at a private healthcare company before moving into big data and data science and into practice as an AI adoption strategist and AI consultant. I hold an MSc in Big Data and Data Science from Northumbria University, awarded with Distinction, for which I developed a conditional latent diffusion model generating synthetic brain MRI for Alzheimer's disease and cognitively normal cases. The work was published in the Journal of Imaging. I have led technical teams in industry, and I am the research and data lead on a project at West London NHS Trust. My research interests lie in the application of AI and machine learning across healthcare, especially medical imaging.

Recent publication, Journal of ImagingRead the paper

Experience

  1. AI Adoption Strategist and AI Consultant

    Independent

    Helping teams adopt AI in their day-to-day work, and designing and building AI agents, MCP integrations and automation.

    More on AI engineering

  2. Data and Research Lead

    West London NHS Trust

    Establishing the research programme for a football and mental-health wellbeing service.

  3. Graduate Student Researcher

    Northumbria University, London

    MSc Big Data & Data Science Technology with Distinction. My dissertation, marked 81, developed a conditional latent diffusion model for synthetic brain MRI in Alzheimer's disease.

    More on the research

  4. IT technician, then IT manager

    A private healthcare company

    Led an IT team of three and a six-person project automating prescription handling.

Research interests

  • Generative modelling for medical image analysis
  • Evaluation methodology for generative models
  • Memorisation and privacy in medical image synthesis
  • Small-cohort learning
  • Neurodegenerative disease

Publications

  1. Conditional Latent Diffusion for Synthetic Brain MRI in Alzheimer's Disease: A Preprocessing-Focused Pipeline

    Fallah, S.; Herzog, N. J.

    Journal of Imaging, 12(9), 426 · Peer-reviewed · Open access · First author · DOI 10.3390/jimaging12090426

    A classifier trained only on synthetic scans reached TSTR AUC 0.779 ± 0.031 against 0.869 ± 0.027 for the same model trained on real data, a difference not distinguishable at this sample size.

  2. Night Interruption with Red and Far-Red Light Optimizes the Phytochemical Composition, Enhances Photosynthetic Efficiency, and Increases Biomass Partitioning in Italian Basil

    Fallah, S.; Aliniaeifard, S.; Zare Mehrjerdi, M.; Mirzaei, S.; Gruda, N. S.

    Plants, 13, 3145 · Peer-reviewed · Open access · First author · DOI 10.3390/plants13223145

Services

Let’s connect

Earlier work in plant science