Transcription of Finance News Network Interview with Volpara Health Technologies Limited (ASX:VHT) Non-Executive Director, Sir Michael Brady.Carolyn Herbert: Hello I’m Carolyn Herbert from the Finance News Network and joining me from Volpara Health Technologies Limited
(ASX:VHT) is Non-Executive Director, Professor Sir Michael Brady. Sir Michael, welcome to FNN.
Sir Michael Brady: Thanks for having me here.
Carolyn Herbert: You’re an expert in imaging analytics. How do the two compliment each other and can you tell us about the latest software, when it comes to breast imaging?
Sir Michael Brady: Images play an extremely important roll in human perception of the world and in particular, for commissions. But they always rely upon the judgment, the experience of the clinician in this particular space. For analytics, we need numbers. We need numbers that represent change. Over time, we need numbers that tell us whether or not how much density there is in a woman’s breast. And what the risk is of that woman developing breast cancer over time.
Now you might say that clinicians, because of their experience, could actually figure out getting the numbers, get the numerical information. The broader evidence is that actually, although they have exquisite skill, they’re extremely poor at deriving numerical information. So we provide decision support to clinicians in the form of numbers computed from images.
Carolyn Herbert: So more specifically, what has this done for medicine?
Sir Michael Brady: So initially, initial attempts to do image analysis in medicine, started out by trying to image anatomy. And then we moved on to imaging physiology and in particular aberrant physiology, which corresponds to illness. And then even more recently, we’ve begun to image function, for example the operation of the human brain. And we’ve begun to do molecular imaging that links together images to cellular processes.
So this has been a huge transformation that’s happened over the past 30-odd years. And we’re able for example to detect features of interest, to identify organs of interest within an image, to detect change over time, to build atlases of normal and abnormal change in organs. So the whole of medicine has been transformed by having information code from images.
Carolyn Herbert: So closer to home, how is Volpara harnessing the power of computing and image analysis?
Sir Michael Brady: One of the first things that Ralph Highnam, who’s the CEO of the company and I did in Oxford way back around 25 years ago, was we showed for the very first time that you could take mammograms, and you could compute very precise numerical information. In particular, we could figure out at every single point in the image, how much fat and how much non-fat tissue or dense tissue, to give it the modern parlance. That turns out to be very closely related to a risk of a woman developing breast cancer, over a five-year period.
It also enables you to stratify those women for whom mammography is absolutely the appropriate imaging method, to determine whether or not they’ve got cancer. Or whether they should have adjunctive imaging, such as ultrasound or MRI. And we’ve moved on beyond that now towards developing the Enterprise System, in which we are supporting the whole of a breast-imaging centre. Both the radiographers, the physicists, the referring physicians, the radiologists and indeed the people who are trying to manage a system that might have hundreds of thousands of women go through, per annum.
Carolyn Herbert: With all data being collected, does the system learn from itself?
Sir Michael Brady: So fundamentally machine learning has had a huge renaissance over the last five years. About fives years ago, if you wanted to try to build algorithms that improved their performance over time, you have to be something of a mathematical hero. But now there are freeware, software systems around and the key now is having large well-curated databases. So for example, our VolparaEnterprise System will probably generate images from around half a million women, just this year alone. And it’s possible to think over the next two/three years that we’ll be getting four/five million more women per annum, which is about five times bigger than the biggest clinical trial ever done on breast cancer.
Carolyn Herbert: Finally Sir Michael. Where do you see medical imaging and analytics evolving over the next decade?
Sir Michael Brady: I think there’s going to be three major trends that we see over the next 10 years or so. The first one is going to be with Cloud delivery of services, which is in fact feeding into machine learning. We will discover information that is latent in the whole complexity of the huge data sources. The second thing is we’ll begin to connect image analysis systems to genetic information. This is a subject, which now goes under the title of Radiomics. But we’re already beginning to discover new things that link together, the feed of typical information from images with the genetic information that’s provided in standard arrays.
And the third thing that we’re doing is we’re beginning to link information, in particular quantitative information, with therapy as illnesses progress over time and respond to therapies. So you can do dose boosting, or you can determine to switch therapies and so forth. That rejoices under the name of Theranostics. So I think the Cloud, Radiomics and Theranostics are going to be the dominant forces, over the next 10 years.
Carolyn Herbert: Sir Michael, thank you for your insights.
Sir Michael Brady: It’s been a pleasure to be here.
Ends
