In databases around the web lie the answers to some of medicine’s most urgent questions but these life-saving discoveries remain hidden within the sheer volume of information generated by advances in scientific research.
“We are kind of drowning in data,” says Professor Sean O’Donoghue, biodata visualisation expert with the Garvan Institute of Medical Research and CSIRO’s Data61, who works to reveal data’s disclosures.
“We’ve already gathered the data and they’re lying buried in databases out there already available, just on the web, but the right people haven’t been actually able to see those data because they’re drowned in the flood of it,” O’Donoghue, whose work is contributing to advances in the study of cancer, says.
“But at some point some human being has to sit in front of those data and analyses and has to see something. Because almost all discoveries are made that way. It’s a person looking at a screen and seeing something they haven’t seen before or they didn’t expect.”
For O’Donoghue, a solution to the problem of penetrating the reams of information lies in data visualisation, which he uses to bring to life complex data sets with the use of images, animation and video that life scientists can use to understand and cure disease.
What I try and do in my work is try to build a system that enables that moment, that moment of insight.
It was in 2000 that O’Donoghue realised moments of insight could be achieved by data visualisation when he was working on a project examining the ground-breaking initial human genome data and analysis that was generated from it.
He set himself a challenge that would become a career-long preoccupation: “which is how do I make all of that mess understandable? Rather than just being a big data dump, how can people actually use the data to gain insight, to gain understanding, to make discoveries? And I’ve been thinking about that problem ever since.”
As a pioneer in the field, he sees it as an area of science in which biology, computer science, animation, art, psychology and design intersect: “to try to keep that control panel sane and tameable.”
The key is understanding how humans perceive data to develop interactive interfaces.
“You are moving through a complex space and doing more complex things, and the interface that a biologist or a healthcare person needs in order to move through this complex landscape of data can get very rapidly overwhelming.”
Working with graphic designers, psychologists, user interface designers and even an Emmy and BAFTA biomedical animator, Dr Drew Berry, has allowed O’Donoghue and his team to produce trailblazing visualisations.
“Ideally you want to give people the freedom to move through the data and to create other uses of the data that makes sense to them, depending on the questions they want to ask.”
With so much data to mine, however, there is a great need for the data visualisation field to develop and collaborate, which is why O’Donoghue co-founded VIZBI in 2010, the oldest and largest international conference bringing together leaders in the field. To be held in Australia for the first time as part of Vivid Sydney’s festival of light, music and ideas from 14-16 June, VIZBI will showcase world-leading visualisations advancing biomedical and biological research. Side events will bring together scientists, artists, innovators and the public.
A goal is to inspire people to join in the effort. As O’Donoghue says, great new discoveries can be had if the data can be effectively mined.
“Building effective visualisations for all these different kinds of data that we have is in many ways one of the brave new steps for discovery nowadays.”
The need is increasingly urgent given the high stakes. At the Garvan, breakthroughs in visualisation are contributing to advances in research into cancer and the structural rearrangements of DNA that occur during the disease.
“Our genome massively changes during cancer, physically, and there are new experiments that have just come online in the last few years that can actually look at how genomes are reorganised in three dimensions, and we can compare cancer to non-cancer. And there had not been good ways to look at those data and we’ve built some new methods.”
The team is also developing “fantastically exciting” techniques in the study of blood that go beyond viewing which proteins are present and in which concentrations to studying each protein and detecting small changes.
“They will enable us to go from a drop of blood to get a detailed picture about what is happening in the person.”
For O’Donoghue the possibilities are infinite: “There’s a lot of work to do.”
Image: 3D Computer Visualisation of the Human Vascular System. Credit: Associate Professor John McGhee, UNSW Australia, 3D Visualisation Aesthetics Lab, https://www.artdesign.unsw.edu.au/3DVAL