Sethu Vijayakumar, Fellow of the RSE, Professor of Robotics, University of Edinburgh
As a mathematician and roboticist, I have been lucky enough to work on many complex projects like developing algorithms for a NASA humanoid robot to balance, navigate and manipulate objects autonomously; in preparation for deployment on Mars.
However, I am fascinated with the question of how the latest advances in my field can help tackle some of the biggest and most intriguing healthcare challenges of our generation.
Personalised medicine and treatment have been the holy grail in medical interventions for a while – with a promise to help reduce collateral damage by being more targeted and effective. One such area of work I have been exposed to is radiotherapy for the treatment of malignant tumours, specifically prostate cancers.
A key challenge in effectively delivering radiotherapy to treat these cancers is ensuring that tissues and organs surrounding the tumour are not exposed to too much harmful radiation. We have been contemplating how robotics can potentially play a significant role in addressing this challenge.
Roboticists know how to program complex robot arms to navigate narrow, complicated spaces in 3D and in real-time, operating even in constantly changing environments. Therefore, instead of using classic (linear accelerator type) radiotherapy delivery devices, which do not have the flexibility to direct rays from all directions, we could exploit multi-joint robot arms to ensure that only the tumour cells are targeted while minimising exposure to the surrounding tissues.
Of course, there are many supporting acts that need to come together to make this a practical reality – for example, real-time imaging, measurement of tumour hypoxia (oxygen levels), adaptive adjustment of the radiotherapy dose, and clinical trials. But it is a challenge worth pursuing since scientists believe that ‘one-in-two people will develop cancer in their lifetime’ (source: NHS website).
Another interesting healthcare domain, and one that is very topical with COVID affecting millions of people around the world, was highlighted in my conversations with practicing clinician and scientist, Professor Kev Dhaliwal. He works across scientific disciplines looking at the treatment of lung diseases (TB, lung injury in intensive care, pneumonia, lung cancer and, recently, the evolving effects of COVID on lungs).
One of the ambitions of a project like this is to deliver personalised solutions by testing various drugs live at the sites of lung inflammation or damage in individuals to chart the best strategy. Sensing, monitoring and delivery of drugs through endoscopic devices is already a challenge, but repeatedly doing this precisely day-after-day to the same site under human guidance is near impossible. This is where automated, autonomous navigation with highly flexible robots can play a significant role, especially if one can use individualistic landmarks inside each patient’s lungs for guidance.
These are but a few examples of expertise in one domain crossing over to address a seemingly complex problem in another – the key message is rather than thinking methods must necessarily become more complex to deliver breakthroughs, we ought to ponder if there are simple solutions to problems from one domain that transgress classical discipline boundaries.
Sethu Vijayakumar is a fellow of the Royal Society of Edinburgh and Professor of Robotics at The University of Edinburgh, Programme Director for Artificial Intelligence at The Alan Turing Institute, and Founding Director of the Edinburgh Centre for Robotics.
This article originally appeared in The Scotsman on Tuesday 02 February 2021.