Alumni Stories
GRADS 2025: James Ireland
“I kept thinking about the pain she must have been in … and, bigger picture, what happens if an instrument is left in a patient post-surgery.”
Reeled in to the research life by the potential for positive community impact, Â鶹´å of Canberra alumnus Dr was looking for a PhD project, when he heard a news story about a woman who had part of a needle left in her body post-surgery.
Retained Surgical Items (RSIs) are medical implements or materials which are unintentionally left behind in patients after surgery – they can cause pain, infection and other complications, often requiring another surgery to remove.
James was finishing up his honours year at the School of Information Technology and Systems (ITS) at UC’s Faculty of Science and Technology, when he learned about RSIs via that news story.
“I did some research on the subject, and I found some studies, but didn’t find that much work being done to circumvent the problem,” he says.
James’ empathy and IT knowledge combined to come up with a solution that used deep learning and computer vision, to count surgical instruments pre- and post-surgery, in a bid to ensure none are left behind.
“To this day, there is a manual process of counting surgical instruments in the operating theatre, to ensure that nothing is left behind in a patient,” he says.
“When you take into account the fatigue that can affect medical staff, as well as the chance of false positives and false negatives – I thought that incorporating machine learning and computer vision into the process might offer a more efficient solution to the problem.”
This week, James accepted his testamur at UC’s September graduation ceremony, with the knowledge that his solution is now closer to becoming a reality – although the pathway he’s taken would probably be a huge surprise to his 13-year-old self.
Growing up in Wamboin, NSW, James’ initial dream was to be a chef.
“I really thought that was what I was going to be,” he says. “When I was 13, I started cooking regularly with my Oma – that’s German for grandmother, and what we lovingly called her. And I really enjoyed cooking.”
A restaurant kitchen apprenticeship changed his mind about that professional path, “… although I still love cooking!” he says.
James was in his mid-20s and considering what he was going to do with the rest of his life, when someone recommended that he study IT.
“It turns out there are some similarities between learning IT and cooking – so much so that cooking analogies feature heavily in introduction to programming coursework!” James says.
He signed up for a Bachelor of Software Engineering (Honours) with UC’s Faculty of Science and Technology, and found the machine/deep learning units intriguing – but the real tipping point for his future came in the form of a winter internship project with the Faculty.
“I worked with [Associate Professors] Wan Li Ma and Dat Tran to explore the potential of using a text-to-speech synthesiser to help people who couldn’t speak, to communicate more easily,” he says.
“I saw the potential for discovery, the problem-solving possibilities, the ability to make a difference for the community,” James says. The more he studied, and the more projects he worked on, the greater James’ confidence grew – both in himself and in his chosen path.
As his honours year drew to a close, James' supervisors Professors Damith Herath – who heads the Collaborative Robotics Lab – and Roland Goecke broached the subject of doing a PhD.
While James’ research mission in safeguarding patients from RSIs was clear, he soon ran into his first challenge: the data on which the machine learning models would be trained.
“The available datasets were not directly related to what we wanted to look at,” he says.
“Ultimately, the biggest research gap lay in open surgery, which was where we could make the most impact – but the datasets available to us were for minimally invasive surgery, so we started with that.”
Unfortunately, researchers from another country presented a paper at an international conference, using the same datasets and with an accuracy rate of 99 per cent for their proposed solution.
“And that meant starting from scratch,” James says. His team concluded that the only way to move forward was to create their own datasets and test if the machine/deep learning models could be trained on synthetic data. James took a six-month stretch to manually annotate the data.
It was all working well, but then came another hurdle: the COVID-19 pandemic. Lockdowns in Canberra meant that James and the research team couldn’t access their labs.
“I remember working till 3am every day, when we could access our labs in between the lockdowns,” James says.
The long days, longer nights and multiple challenges took a toll, and James reached the point of burnout.
“It was a very low part of my life, and I found it hard to properly function at that point,” he says.
James wisely took a six-month break to recover – he credits his supervisors for their understanding and support.
“I saw a psychologist and read a lot of self-help books,” he says. And he spent a lot of time at art galleries.
“The worst part of the burnout was feeling like I had lost my creative spark as well as my motivation,” James says. Visiting galleries allowed him to enjoy the artists’ creative approaches and process, and to learn other people’s stories of overcoming.
It was a slow ascent out of the slump, but James was quietly determined … and made it through to continue his PhD journey.
“I learned a lot from the burnout. I was initially very excitable and impulsive when I first started out – there was a point that I was working almost 24/7. But I know now the importance of taking breaks, pacing myself, and being able to clock off,” he says.
While the recovery time had slightly delayed his first publication, the second and third contributions were hot on its heels. And most importantly, James’ studies validated the benefits of the synthetic data used to learn counting on real world images.
“My third contribution focused on having multiple cameras in an operating room – you have to consider optimal placement, and also take into account that you’ll have multiple people walking around,” he says.
James may have completed his PhD, but his research journey will continue.
“We now need a user study for nurses and doctors, to see what works best for them,” he says.
This next step is in the works, as his team applies for ethics approval to work with hospitals to collect surgical operational imagery. “It’s a complex issue, as we are looking to navigate privacy issues, possibly by taking very close-up photos of the instruments, so it will take time – but it is work worth doing,” he says.
Throughout his PhD journey, James’ research has consumed so much headspace and heart-space – but beyond the research outcomes, he also knows that navigating his way though burnout helped him grow as a person.
“When I’ve loved something, I’ve always been fully into it – and that made my burnout so much harder,” he says.
“But taking the time to recover helped me rediscover my passion for my work, even if I am a bit more battle-weary now. Doing this PhD has been such a rewarding life experience … both when things worked out, and when they didn’t.
“Nowadays, when I’m frustrated, I sometimes just look at my thesis. Seeing what was possible – when it seemed impossible – cheers me up.”
Words by Suzanne Lazaroo, photos by Tyler Cherry and UCMedia team.
This September, we celebrate the graduating class of 2025.
Our graduates are from many different places, at various stages of life, with diverse skills and perspectives. In their time at UC, each one has contributed to the fabric of the Â鶹´å community. While their journey here may be complete for now, their actions will leave an indelible imprint on generations to come.
We congratulate the graduating class of 2025; we can’t wait to see where you will go next.