The Future of Endometriosis Imaging and Diagnosis: A Conversation with Dr. Jodie Avery
Interview with Dr. Jodie Avery is a Senior Research Fellow in the endometriosis research group at theRobinson Research Institute at The University of Adelaide.
She has a keen interest in women’s health and has developed studies to improve diagnosis of endometriosis using imaging testsand taking advantage of artificial intelligence.
TWEM believes that raising awareness and advancing research go hand in hand - real change comes from shining a light on the work driving progress . We sat down with Dr. Jodie Avery to learn how her research is pushing endometriosis diagnosis forward.
Q: Tell us about your experience as a researcher and what brought you to study endometriosis.
Dr. Avery: I'm part of a team led by Professor Louise Hull who is a fertility specialist and a gynaecologist in Adelaide. She found a lot of her patients who were coming for fertility treatment also had endometriosis but they hadn't been properly investigated. She would send her patients off for imaging and the imaging that she would receive back was not of great quality. When she requested specific transvaginal endometriosis ultrasounds, she would not be getting these ultrasounds because it turns out that not many sonographers actually know how to do these scans properly. So, we thought, how are we going to make this better? How are we going to make sure that women do get a diagnosis without having to go through surgery? So, she thought, maybe if I combine specialist endometriosis ultrasounds and MRIs together, we can see that there's going to be better diagnostic accuracy in these tests. She also had evidence to back this up because she had undertaken five Cochrane systematic reviews on all different biomarkers of endometriosis, including imaging, blood, urinary, endometrium, and the combination of a couple of those ones. And she found that the best way to diagnose endo non-invasively was to use these imaging tests. So we brought a team together using gynecologists, specialist sonographers, radiologists, and AI scientists from the Australian Institute of Machine Learning in Adelaide. We worked out a way of collecting lots of imaging from people who'd already had imaging. We've been developing an algorithm and we've got some really excellent results, and managed to get a patent as well. So we just continue on with that work and we also will be validating that algorithm on a prospective study of almost 400 women as well.
Q: Are these participants going to be all from Australia for now or is the validation going to be worldwide?
Dr.Avery: We started off with Australian data. We also have a very good friend who's part of our study called Dr. Matthew Leonardi, who's Canadian, and he's supplying us with a whole lot of ultrasounds from his practice. We're also looking at expanding into the US and the UK because, in order to commercialise our product - which will be a product that we can give to diagnostic radiography practices to embed in their PACS system to help the radiologists with these diagnoses - we need to get data from all over the world so that we can eliminate any of the biases that come about when using AI. That's a big task for this year. So we've just got a bit of money to help us do that.
Q: Currently your studies in your lab have been focusing on using data from these ultrasounds that have been shared with you and your team. Are you developing any other studies with clinical patients at the moment?
Dr.Avery: I've received another couple of million dollars to look at about 400 adolescent girls within six months of like the beginning of their diagnostic journey. They can't have had surgery yet because we want to see what their trajectory is and how an imaging diagnosis affects their quality of life.I’ve got a PhD student really concentrating on the quality of life of girls with endo.
At the same time, we're trying to develop a new quality of life tool called the HERA-QoL-P. We've actually started developing this in PCOS, but we're bringing it into endo as well.
Another project looks at family physicians’ barriers to diagnosing endometriosis. It's really quite hard to get an appointment with a gynaecologist in Australia and there aren't enough gynaecologists to go around to do surgery on all the one in seven women that have endometriosis. So we're trying to bring it back to the basic level to get women the support they need initially before they have to go through specialist appointments.
We've got a website called EndoZone, it’s a one-stop shop for women with endometriosis just to find out about management, different therapies, and different research they can be involved in. Lots of really great information and it's all evidence based.
We've got a master's student looking at accessibility of a diagnosis in Australia because just like Canada, Australia is a pretty big country and very remote in some areas. We want to make sure, especially with our First Nations people, to make access to diagnosis better because some live so far away and they don't have access to a big MRI machine. We'd love to be able to get them a proper endometriosis scan and then get someone, say remotely, digitally to be able to read those scans for them.
Q: Something that caught us by surprise was when we were analyzing our metrics on our social media, we realized a big part of our audience comes from Australia. It seems that Australia is advancing a lot in terms of research funding for endometriosis. We also see nonprofits being created and advocating for people suffering from this condition, as well as some public politics that have been changed recently. What is your perspective on this matter? Do you see Australia advancing or is this just an outsider perspective?
Dr.Avery: About five years ago, the Australian government had, I think, the Parliamentary Friends of Endometriosis and they started the National Action Plan on Endometriosis.The Prime Minister at the time, his wife had endometriosis and it was a bit of running joke because he wasn't a very popular Prime Minister but he used to ask his wife everything. So if there was ever a big crisis, he'd go, “What would Jenny think?”. It turned out that Jenny had endometriosis. So we think that that might have been the start of something. There were a few female politicians also who suffered from endometriosis who were very vocal. A lot of people were also lobbying for money as well and there are lots of different patient advocacy groups in Australia and suddenly some money became available.
But then they sort of formalized all the advocacy group networks. So now there's this big group, there's Endometriosis Australia, but then there's also the Australian Coalition of Endometriosis and a big group called QENDO (Queensland Endometriosis). In South Australia, one of the politicians herself had endo and after talking to us one day, she decided to start a thing called a select committee on endometriosis as well. A lot of different people presented on what had happened to them and their stories and what we could do to change things. That happened last year.
One of the really great things was that the government funded 22 endometriosis and pelvic pain clinics across the country. They're going to have another 10 more funded as well. So that's a really, really great initiative because it just brings a one-stop shop for girls with endo. So a lot of the politicians are becoming very aware and the lobby groups are huge. There was lots of stuff in the media as well, so I just think people are finally listening to young women in this space and that's why attention is getting in this area more than it used to be. It's a great time to work in the area as well because of all this support and everything that's happening.
Q: We have also seen movement in the biotech companies regarding developing tools, kits, tests to diagnose individuals. In your lab, you have been developing imaging tests and improving them using AI to help in this matter. But we also see other biotech companies using saliva, endometrial tissue, blood, even menstrual blood to try to detect a biomarker that can be used to diagnose endometriosis. How do you see this development? What do you think will be the future of diagnosis? And more importantly, what do you think is going to be the role of AI in helping to diagnose endometriosis?
Dr.Avery: A lot of these tests, specifically the blood and the salivary tests, all they really test is inflammation. If someone's got inflammation in the body it could be arthritis, it could be anything really. It's a tool to say, yes, you have endo, but you really need to come in and do imaging afterwards to say this is where it is. If someone just does this test and they say, you've got endo, you need surgery, the surgeons aren't necessarily going to be prepared to know where the endo is. If the endo is on the bowel, you need to bring a colorectal surgeon in to assist with that operation. But if you don't know where the endometriosis is without using imaging, then you've wasted time because the doctor will open the women up and go, “This is too hard, you'll have to come back later when we're better prepared” and shut them. So that means there's gonna be repeat surgeries and we don't want that to happen to anyone. So yes, these tests are good as a screening tool, and a lot of money, especially in France, has been put into these tests which haven't really come up with really great results yet. It remains to be seen what's going to happen. With our tests we've been getting quite high sensitivity and specificity. We will be able to tell where the endometriosis is specifically with Pouch of Douglas obliteration, bowel, and nodules. We're working on endometrioma and uterus sacral ligaments at the moment, and we will be including more signs in the future. We will also possibly be including other modalities like nuclear medicine to help look for things like superficial endometriosis, which has been quite hard to find in the past.
We're constantly developing our program. We are always working with our AI team. We work with Professor Gustavo Caniero. He's one of the world experts in computer vision. We also have Dr. Yuan Zhang, and she's very primarily responsible for developing the algorithms. So we keep constantly updating these things. We'll be using things like federated learning as well so that we can get over privacy issues in transferring data from one country to another. The main problems we have are that we do need lots and lots of data. In endometriosis, people aren't screened. It's only when you've got a problem that you will go and have a scan. So that's our main problem, not being able to get enough data, but we're trying very hard. Every day something new with AI happens. So we've got to keep up with that kind of thing and we've got to make sure we've got the right people who can do that kind of thing for us as well.
Q: You mentioned most of your data you want to expand to the US and the UK. Are there any future plans to expand to other developing/non-western countries (to represent more people of colour) in places where there are high rates of endometriosis?
Dr.Avery: Yes, we have been talking to our collaborators in Indonesia and we will try, but each time we go into a new country, we have to get through ethics and get the right collaborators and things like that. So it does take a lot of time. There's a huge endometriosis movement in India as well. So, hopefully we'll be able to tap into that. And China. So we will continue to try and get into other populations as well. And I think luckily, in Australia, Canada, America, we're all pretty multicultural. So we're able to get lots of different people in our data sets, but the make-ups are slightly different across. And apparently data from other countries isn't quite as good as data from America. If we ever want to go down that route to get FDA approval, that's going to be interesting. So we're trying.
Q: Do you have any final message for our audience?
Dr.Avery: I think in the future, there's so much hope for women with endo. There are new treatments that are coming on board as well. Like there's a great new drug called Relugolix, which is coming out in Australia, which is showing some huge promise. Drugs haven't been developed for a very long time and suddenly there are lots of new drugs out there so hopefully that will be a great thing. We actually need to make sure we get proper diagnosis first before we jump into that and make sure that we can support women through going to school and making life much better for them.
