“When Spring comes to Paris the humblest mortal alive must feel that he dwells in paradise” - Henry Miller
I arrived in Paris on Sunday, March 26th, 2018. The sky was grey, the train was late, and the platform smelt of bodily fluids and old train. It didn’t feel like paradise to me.
This was the first day of my PhD at the CEA (French Atomic Energy Commission) in Paris, where I would be working for the next three years. The project is part of the HYBRID2020 programme, which, according to the website: ‘brings together international academic, industrial and non-governmental organizations (NGO) partners in a cross-speciality platform to advance imaging sciences as part of personalised medicine with a focus on molecular and hybrid, anato-metabolic imaging’. I have no idea what this means either. My project would be focussing on analysing cancer-based medical images (MRI, PET, etc.), the aim being to use machine learning techniques to identify aspects of the scans that could indicate the best treatment for the patient.
What compelled me to apply for a PhD in France, a country where I have to communicate via hand gestures and Google translate? The food and wine were certainly factors, but a big consideration was the HYBRID programme. Of course I’m in a world-class research institution working with great people, but what sets HYBRID apart is the breadth of support in other areas. With a network of other PhD students I can bounce ideas off people working on similar problems. Funding for training, conferences, summer schools and secondments mean I can connect to the wider medical imaging community, an invaluable asset during my PhD and for future career options. I’ve also had the chance to take language classes, so I can fully assimilate into ‘la vie Parisienne’.
“To err is human. To loaf is Parisian” - Victor Hugo
August in Paris is wonderful. By the Seine students eat myriad cheeses and drink those very-French mini beers and discuss Voltaire into the night (I like to imagine). In the parks old men bicker and laugh over afternoon games of pétanque. In the bistros coffee cups clat, women chat, and the whole city lets out a sigh of contentment as it soaks up the sun.
Away from the balmy allure of the Parisian parks, in the lab my work is getting into its stride. I recently defeated the Kafkaesque French banks and managed to open an account, and am now fully focussed on my research. The first step was to immerse myself in my topic. This meant finding a comfy chair and reading a big wad of scientific papers. I also needed to get up to speed on the technical side, having little prior experience in machine learning, especially in a medical imaging context. To this end, I have been experimenting with simple convolutional neural networks to categorise 2D PET lung images as ‘cancer’ or ‘no cancer’. It’s not going to get me on the front page of Nature, but it does help understand the concepts of deep learning and the tools I will be using. To prevent me from drowning in an ocean of new concepts and terminology, I get thrown life-jackets in the form of regular meetings with my supervisor, Irène Buvat, to discuss ideas, difficulties, and advances in the medical imaging community.
But I haven’t just been staring at my screen wondering what a Heap Error is. In July I had the opportunity to go to ‘TOPIMTECH 2018’, a summer workshop in the beautiful fishing village of Chania, Crete. This was a week of talks and practical demonstrations on the topic of ‘Big Data in Imaging’. It was my first experience of a conference-style trip, and a great thing to do so early into my PhD. The range of people (mathematicians, physicists, computer scientists, biologists) meant I got input from completely different viewpoints, and when I left my brain was fizzing with new ideas to try out.
Written by David Wallis