Work Experience Students’ Diaries

Posted on By Outreach
Categories: General
Tags:

This summer, two work experience students joined VISIONLab to learn about our work, and experience working in a multidisciplinary research lab.

Antara Roy and Rory Dunbar met with members of the lab to learn about the varied projects we undertake here at the Cavendish Laboratory.

As part of this experience, Antara and Rory wrote the pieces below, describing their time in the lab, and explaining some of the concepts they learned.

We would like to thank Antara and Rory for their enthusiasm and energy during their stay, and for engaging eagerly with all the researchers they met. Good luck in the future!


I arrived at the Cavendish Laboratory on the 2ndJuly and experienced 1 week at the Biological and Soft Systems group. It was a wonderful experience that gave me great insight into the experimental work of professional scientists. In VISIONLab, a team of scientists aim to develop new and more sensitive imaging modalities to detect the early onset of cancer.

I visited the nanophotonic lab with Sophia and the endoscope development lab with Dale. I explored phase imaging with Abby and was introduced to fascinating Raman spectroscopy by Ben. I saw the production of phantom tissue with James, the use of holographic imaging with George, and learned about the complex, yet wonderful, world of machine learning with Alex.

Ben explained that he is exploring the link between oxygen and tumour development. By shining coherent light (for example, laser light) on cells, and measuring light that is inelastically scattered, (scattering of light where energy is conserved), Ben probes the vibrations of molecular bonds (the bonds of molecules within the cell).

This technique, called Raman spectroscopy, allows us to detect the type of molecular bond by matching the energy detected to the known energy of different molecular bonds present in the cell. This allows scientists to detect the different chemical compositions of cells such as lipid, proteins and DNA, and track how these change as cancer progresses. This work is still far from clinical application, but research so far proves to be promising.

Many members of VISIONLab are focussing on oesophageal cancer. Oesophageal cancer is the cancer of the food pipe. It is hard to treat in the later stages, so the five-year survival rate is as low as 19% [1]. As a result, VISIONLab are working on new innovative technologies that can be used to aid early diagnosis. These new technologies, have the potential to reduce the fatality rate considerably by identifying cancer in its early stages, so that it can be treated before it spreads.

Currently, diagnosis of early cancer is performed with endoscopes which image using white light. However, cancerous tissue and non-cancerous tissue is difficult to distinguish, resulting in a high miss rate. New imaging techniques, which allow better identification of early cancer, are therefore integral to improving early detection.

Jonghee and Dale work on one of these new techniques, Hyperspectral Imaging (HSI). This method of imaging collects and processes many wavelengths from across the electromagnetic spectrum. Humans can only see light from three wavelength ranges or ‘bands’ of the spectrum; red, green and blue.

Hyperspectral imaging detects 10s to 100s of bands from across the spectrum, and not just from the visible region, but from the near infrared region too. Furthermore, HSI allows us to detect smaller bands of wavelengths, around 10 – 20 nanometres wide. The vast amount of information gathered by HSI has the potential to allow the detection of biological molecules within early cancer, for example metabolic molecules such as NAD(H) and FAD(H), based on their unique spectra. The detection of these molecules could lead to more accurate detection of early cancer.

To analyse the complex images generated by HSI, Alexandru is working on a method know as machine learning. This complex computational technique involves using man-made neural networks (mimicking the use of neurones in the brain). We can ‘teach’ this neural network computer to differentiate between different images, for example, early cancer and non-cancer. In the future, this technique could be used to sift through endoscopy images with greater speed and accuracy than a human interpreter, improving early diagnosis.

This was perhaps my favourite project. It baffled me to think of an algorithm that could result in a computer being able to make ‘decisions’. I was also taken aback by the technical equipment available in the labs – from atomic force microscopes to lasers. I feel as though I was introduced to world class experimental labs.

My work experience introduced me to the working life of real scientists, not the ones in movies or The Big Bang Theory. I was introduced to a world of conferences, paperwork, and last but not least, amazing experiments! It gave me an idea of what a scientific community looks like – and opportunities lie ahead – physics is definitely for me.

I would like to thank all the people who worked with us during the week, Dale for organising everything and Sophia, Abby, Ben, George, James, Alex and Lina for showing us around their labs and answering our burning questions. Without their time, this would not have been possible.

Thank you!

Antara Roy


Since my arrival at Cavendish Laboratory, I’ve had an absolutely amazing time! Everyone has been really kind and shared their knowledge with me and throughout my whole week I’ve remained in a constant state of fascination. After each piece of information has been explained to me, I’ve spent my time researching what I’d learned to confirm my notes were accurate and to develop a deeper understanding.

First we learned about a technology called multispectral endoscopy. Dale, who supervised me during my week of work experience, has developed a multispectral endoscope which is currently in clinical trials for early detection of cancer in patients with Barrett’s oesophagus. This device should allow doctors to more effectively detect early signs of cancer called dysplasia, allowing it to be treated before it develops into the more dangerous adenocarcinoma.

When using white light imaging, it is often difficult to observe differences between colours of cancerous and non-cancerous tissues because we only detect 3 colours; red, green and blue.

However, multispectral imaging, which is used in Dale’s multispectral endoscope, can be used to discover spectral differences by detecting a wider range and wider number of colours. Using these extra wavelengths and spectral range has the potential to help us spot subtle differences between early cancer and non-cancerous tissue that are not visible in white light imaging.

Perhaps my favourite experience this week was learning about Alex’s work on machine learning systems. These will be designed to automatically interpret large multidimensional data sets generated by Dale and Jonghee using their multispectral endoscopes. It was particularly fascinating for me to me to see how machine learning technology could be applied to a clinical problem to distinguish between images of different disease types.

I also enjoyed learning about optoacoustic tomography from James and Lina as theirs were perhaps the projects I best understood.

In optoacoustic tomography, non-ionising laser pulses are delivered to biological tissues. Some of their energy is absorbed and converted into heat, leading to a transient thermoelastic expansion and ultrasonic emission. The ultrasonic waves, commonly called ultrasound, are detected by ultrasonic transducers and analysed by a computer to produce images of the tissue.

When optoacoustic tomography is performed at different wavelengths, a technique called Multi Spectral Optoacoustic Tomography (MSOT), we can use the variation ofsignal with wavelength to assess the vascular function of tissue without invasive procedures.

I loved all the projects as each was interesting in its own unique ways. Probably the most fascinating concept I encountered was Machine Learning, as it was incredibly fascinating to hear about and was expertly explained by Alex. The concept of machines effectively teaching themselves to operate was incredible and it is awe-inspiring to think of all of its future applications when the software is fully developed and understood.

I would like to sincerely thank everyone who I worked with as they were all incredibly understanding and kind. I would love to continue down the physics route towards university and would give serious consideration to learning more or continuing down this specific field of science. Also, I am more motivated than ever to do my best to return to the Cavendish if all goes well in my next couple years of studying. All in all, I don’t think I could’ve been any happier after this week and would love to thank everyone that helped make this experience possible!

Rory Dunbar


[1] Lao-Sirieix P, Fitzgerald R. Screening for oesophageal cancer. Nat Rev Clin Oncol 2012;9:278–87.