Rob Brisk, MBBCh, MRCP
Speciality doctor in cardiology, Clinical Research Fellow
 

It’s an exciting time to be both a practising clinician and a deep learning engineer. Deep learning is allowing us to do incredible things with complex data, while modern medicine is overflowing with data-rich problems: it’s a match made in heaven. Yet, the two worlds have never been more subspecialised nor so disparate, and linking up the clinical problems with the technological solutions has been a huge stumbling block for many promising digital health companies.

As someone who has spent most of their waking adult life as a doctor on the front lines of acute hospital medicine, I am more at home in the clinical environment than anywhere else. I understand the pressing needs of modern adult medicine and the complex workflows within which they exist. But as someone who has also spent the last decade studying computer science, who has designed, developed and deployed both conventional and deep learning software solutions to real-world clinical problems, I’m uniquely placed to see how digital health solutions could best fit into that environment.

We’re at the crest of a wave of incredible, AI-powered technology that could change the face of healthcare as we know it. The million dollar question is whether that technology can find its way into the hands of those who need it the most. I believe that those of us who are as comfortable writing neural networks in Python as we are in the clinical consulting room have an invaluable role to play in this process, and I’m excited to hear about any innovative digital health projects that might benefit from a clinician-programmer’s perspective.