Last week, Sanome’s Machine Learning Lead Dr Andrew Creagh, and Data Scientist Tom Pease, attended the five-day conference hosted by the Cambridge Centre for AI in Medicine (CCAIM).
Tom Pease reviews the week below.
The Cambridge Centre for AI in Medicine (CCAIM) Summer School was a brilliant week – a chance to hear from some of the great and good of the medical AI world, and take the pulse of the field at large.
The conference opened with Dr Eoin McKinney speaking about the ‘AI Chasm’ – the current gap between the potential of AI in healthcare, and the fruits of that potential. He emphasised that, despite AI products maturing rapidly in the world of commerce, in the field of Healthcare, AI is still at the stage of future potential. IF this was a diagnosis, it’s clear that the Van der Schaar lab have been working on creating the tools necessary to build the cure — alongside the very fact of setting up the conference, during the week there were talks and demos on practical tools that have been developed by researchers at and adjacent to the lab. These included SimplEx, an out-of-the-box tool to help researchers make their models interpretable and therefore more practically useful, and AutoPrognosis, an ‘AutoML’ package designed to automate the process of finding the best model for a problem, therefore freeing up time and energy for less automatable tasks.
Alongside the new ideas for processes to follow and tools to use, there were also some fascinating talks by speakers detailing AI projects that have been deployed or are being trialled in clinical settings. Particularly inspiring was Professor Andres Floto, talking about Project Breathe, which uses machine learning to identify health deteriorations in cystic fibrosis patients significantly earlier than they would otherwise be detected, and enables clinicians to take actions to prevent or ameliorate the deterioration.