Infrared Imaging Techniques for the Diagnosis and Characterisation of Cancer

Cancer is a leading cause of death, with 8.2 million cancer related mortalities worldwide in 2012 [1]. Whilst treatment is the primary target of funding for cancer research, early diagnosis is an essential field for development. A Cancer Research UK (CRUK) study showed that 90% of bowel cancer patients survive longer than five years if the disease is diagnosed at the earliest stage [2]. Histopathological examination of disease tissue is the gold standard for cancer diagnosis, as it is able to extract a large amount of morphological and architectural information about the disease state. Despite the success attributed to histology, there are significant drawbacks. Inter observer variability in the diagnosis of low-grade disease is amongst the most important, as incorrect diagnoses have the potential to lead to unnecessary high-risk treatment, and in extreme cases - death. It is also a time and resource intensive field, as highly trained professionals are required to stain and examine a large quantity of samples. There is therefore a demand for automated, accurate and high throughput techniques to augment clinicians in diagnostics.
Infrared (IR) imaging is a well-established technique for spatially resolving materials as a function of chemical identity. There has been considerable focus from research groups of a variety of disciplines in applying the strengths of IR imaging to the characterisation and diagnosis of disease, especially cancer, and with the aid of advanced technology and analytical techniques, this field has been able to progress rapidly in recent years. The high throughput nature of IR imaging makes a convincing case for clinical use, however there is still a wide gap between demonstrated potential and application.
Whilst the work on this project is open ended, there are two key objectives. One is to establish several models to discriminate between given pathologies. This will initially be using a new, novel multivariate algorithm which has shown great promise in preliminary studies. Once a suitable model has been established, it will be tested against unknown data to assess its performance, and to evaluate its feasibility as a potential diagnostic tool. The other objective is to exploit different IR imaging techniques to explore the chemical and mechanical properties of tissues, in an effort to develop understanding of the disease state.

References

1. [1] L. A. Torre, F. Bray, R. L. Siegel, J. Ferlay, J. Lortet-tieulent, and A. Jemal, "Global Cancer Statistics, 2012," CA: a cancer journal of clinicians., vol. 65, no. 2, pp. 87-108, 2015. [Online]. Available: http://onlinelibrary.wiley.com/doi/10.3322/caac.21262/abstract
2. [2] CRUK, "Bowel cancer statistics," Cancer Research UK, 2017.