Overview
This Theme will strengthen the UK response in the event of an accidental or deliberate ionising radiation release. We will work with Theme II to design robust blood-based exposure/effect biomarker studies, using new technologies including AI, to assess who has been exposed and at what dose. We will plan and conduct follow-up studies of affected (and non-/less-affected) individuals, including exposure/effect biomarkers to provide an integrated assessment of medium- to long-term health effects. Alongside this, we will undertake health surveillance at the small-area scale leveraging the Small-area modelling (SAHSU) resource using routinely collected health data, as we have done in our surveillance work for Committee on the Medical Aspects of Radiation Exposure (COMARE) around nuclear installations. As well as the HPRU in Emergency Preparedness and Response (HPRU-EPR), we will work closely with the HPRU in Evaluation and Behavioural Science (HPRU-EBS) on all projects within this Theme.
Projects
- Project 1a. Development of biodosimetry, including AI - Developing machine learning models for radiation dose prediction
- Project 1b. Development of biodosimetry, including AI - Machine learning, artificial intelligence and automation to improve biodosimetry assays
- Project 2. The effect(s) of radiation quality and (high) dose rate exposures during radiation releases from unplanned nuclear events
- Project 3. Long-term follow-up - Developing evidence-based approaches to population health monitoring and communication
- Project 4. Nuclear installations and childhood cancer
Project Leads: Philip Davies (KCL), Sophia Tsoka (KCL)
Project Team: Modest Lungociu (PhD student, UKHSA/KCL)
Rapid and accurate estimation of radiation exposure is necessary to effectively triage casualties in emergency situations such as nuclear reactor disasters and atomic bomb explosions. Current methods for radiation biodosimetry rely on assays which can be time-consuming, lack scalability, and may not adequately capture the full biological impact of radiation exposure across diverse individuals.
Recent advances in DNA and RNA sequencing technology, particularly with third-generation platforms like Oxford Nanopore Technologies (ONT), enable large-scale generation of biological data from minimal blood or skin samples. Specifically, the cellular state of an individual can be assessed by measuring gene expression levels, detecting DNA mutations, identifying RNA modifications and RNA splicing perturbations from a single sample. This unprecedented breadth of information promises to revolutionise personalised medicine, with huge implications for radiation protection.
The aim of this project is to develop the next generation of models used for radiation exposure dose prediction, by leveraging cutting-edge machine learning (ML) methods together with “long-read” sequencing data derived from laboratory experiments and real-world samples of individuals exposed to radiation. The use of other covariates, such as time elapsed since radiation exposure, gender, age or infection status will also be explored.
Project Leads: Hannah Mancey (UKHSA), Dragana Vuckovic (Imperial)
Project Team: Stephen Barnard (UKHSA), Marc Chadeau-Hyam (Imperial)
Biodosimetry assays enable the estimation of the radiation dose received by a person using biological material, typically a blood sample. Currently, the bottleneck in this process is the requirement for trained experts to score large numbers of individual cells to give a single dose estimation, meaning providing an accurate dose estimate can take up to 5 days. In the event of a mass casualty radiation exposure, these dose estimations would be critical for triage and treatment, ensuring that appropriate medical care could be provided to those affected, with priority given to individuals who have received the highest doses.
The primary objective of this project is to identify and implement tools and approaches that increase the capacity for dose estimation in biological endpoints during mass casualty radiation exposures. This is part of project 1, which focuses on the development of biodosimetry, including the integration of artificial intelligence (AI).
AI and machine learning tools could be employed for automated cell scoring to identify radiation damage. Given the high accuracy and reliability required, these tools will likely need to be trained extensively using large datasets. This project aims to identify a tool that can be adopted across the biodosimetry community by collaborating with international biodosimetry response networks to share relevant data.
Project Leads: Mustafa Najim (UKHSA), Danny Fresstone (UKHSA)
Project Team: Yannick Comoglio (UKHSA), Anil Gunesh (Imperial)
In the situation of a nuclear incident, it’s crucial to have a quick and high throughput methodology to screen individuals in the population who are not wearing a physical dosimeter and who may have been exposed to radiation. This is required to be able to provide urgent medical attention to the exposed ones. Gene expression and cytogenetics methods have proven to be good biological indicators of radiation exposure, helping to estimate the dose received. In particular, gene expression analysis in circulating white blood cells and skin can provide rapid and accurate dose estimates to triage large numbers of individuals.
During a nuclear event, various types of ionizing radiation are emitted, and they can be received at different dose rates based on proximity to the explosion. Different radiation qualities differ in their mass, charge and energy and consequently its ability to cause biological or physical effects. Each type of ionizing radiation (e. g. alpha particles, electrons, protons, neutrons, heavy ions, X-rays, gamma rays) has different relative biological effectiveness. For instance, neutrons are roughly ten times more effective at causing biological damage compared to gamma or beta radiation of equivalent energy exposure.
Hence, the objectives of this project include the development of protocols for the use of robust blood and skin-based exposure markers and the characterization of gene expression-based biomarkers as well as cytogenetic methods to different radiation qualities and dose rates (incl. FLASH), linked to nuclear events.
Project Leads: Bethan Davies (Imperial), Liz Ainsbury (UKHSA)
Project Team: Richard Haylock (UKHSA), Stephen Barnard (UKHSA), Matthew Simpson (UKHSA), Lauren Ashworth-Donn (UKHSA), Aimee Harragan (UKHSA), Holly Carter (UKHSA), Katy Turner (UKHSA), Richard Amlot (UKHSA)
As a category 1 emergency responder, UKHSA has responsibility for the public health response to a radiation accident or incident. Health surveillance and reassurance monitoring following a small or large scale event is essential to facilitate long term recovery. The aim of this work will be to strengthen the UK emergency preparedness and response plans in response of population health monitoring. UKHSA has direct experience of population based follow up to such incidents, but more work is needed to embed appropriate small-area scale epidemiological approaches (Imperial’s area of expertise), and further engagement with stakeholders including the public is also needed, to ensure readiness to respond appropriately and effectively. The Project will be run in collaboration with the HPRUs in EPR and EBS, with input from collaborators at WHO and other international partners. The work will focus on a review of current systems followed by broad UK and international stakeholder input to design new protocols for long-term health surveillance, and related communication and engagement strategies pre- and post-accident, including for those with related health anxiety who may well represent a disproportionately large burden on the health system during the recovery phase.
Project Leads: Bethan Davies and (Imperial), Antony Bexon (UKHSA)
Project Team: Daniela Fecht (Imperial), Fred Piel (Imperial), Monica Pirani (Imperial), Kelly Jones (UKHSA), Mireille Toledano (Imperial), Paul Elliott (Imperial), Jayne Moquet (UKHSA), Anil Gunesh (Imperial)
The UK government is committed to nuclear power with the ambition to increase nuclear energy capacity to support Net Zero and energy security agenda. In the first instance, this is estimated to require 8 new reactors (in addition to the 9 currently operational on 5 sites). Small modular reactors and fusion are also areas of key strategic interest, and there is a continued need to identify appropriate disposal methods and sites. Concern about the health impacts of residential proximity to nuclear installations is longlasting.
In the first year, we propose to establish further longer-term large-scale population surveillance studies around installations in relation to childhood cancer, and to determine baseline risks of the disease in populations in proximity to proposed nuclear development sites.
