Applying an astrophysics modelling tool to improve the diagnosis and treatment of cancers

The use of light is fundamental to a diverse range of medical advances in diagnostics and therapy. For example we can use a handheld devices to detect elevated bilirubin levels in neonates, apply lasers to surgically resect tissue, deploy cameras to image cancerous tissue during surgery, and destroy cancer cells using photodynamic therapy. In the future we may be able to treat tumours using optically heated nanoparticles, or diagnose breast cancer without the necessity for a biopsy.

All these technologies rely on a deep understanding of the complex interplay between optical radiation and tissue. As light propagates through the body it can be scattered and absorbed, it can cross boundaries between different tissues where it can be reflected and refracted, and its frequency can be changed by inelastic scattering or fluorescence. Furthermore the material through which the light travels may be highly heterogeneous and anisotropic, and display intricate anatomical structures on a wide variety of scales.

If we can develop a numerical model of sufficient complexity we may construct a digital phantom of the relevant tissue, and conduct numerical experiments in silico. We can fabricate model devices, illuminate tissue with a wide range of intensities and wavelengths, and investigate how varying the model parameters affects the heating of the tissue, the dosage for photodynamic therapy, or the detection of fluorescent photons. Furthermore these experiments can be conducted rapidly, cheaply, and safely.

Despite the vast differences in scale, the physical and numerical challenges that scientists face when modelling radiative transfer (RT) in the human body are in essence the same as those faced when computing RT models of stars and galaxies. Both regimes contain highly scattering media with complex morphologies on a large range of scales, where polychromatic radiation must be considered and where the absorption and scattering coefficients change with environment. Critically the use of Monte Carlo (MC) modelling techniques (in which light is modelled via a large number of photon 'packets') has become the gold standard in both fields. It is therefore natural to try and exploit the algorithmic and computer science advances that have been made in astrophysical MC codes and apply them to biomedical physics, and this is the essence of this proposal.

The University of Exeter is home to the TORUS Monte Carlo code that has been used to model a wide variety of astrophysical phenomena, from planet formation in discs to 21-cm radiation in galaxies. Harries and his team have developed TORUS for over 15 years. In 2016 we obtained funding from the Wellcome Trust to adapt TORUS to tackle biomedical problems. Subsequently we have been awarded a 4-year Carlotta Palmer PhD studentship to investigate the optimisation of photodynamic therapy (PDT) in the treatment of non-melanoma skin cancer, and a 3.5-year EPSRC DTP studentship to model Raman scattering in breast tissue as an aid to cancer diagnosis.

We wish to hire an STFC Innovation Fellow to join our nascent biomedical modelling team. The Fellow will work alongside physicists, mathematicians, biologists and clinicians in applying our modelling tool to the diagnosis and treatment of diseases such as breast and skin cancer. They will also work closely with our technology transfer office in investigating the commercial potential of our code by engaging with companies and other non-academic organisations. Throughout the fellowship we will provide the training and mentoring needed to accelerate the Fellow's career to the point at which they can successfully apply for independent funding.