Lumina Nurtures AI Innovation to Transform Healthcare

News - 09 Jun 2022

Queensland’s Gold Coast is positioned to become a leading global hub for MedTech, supported by strong capabilities and innovation in artificial intelligence (AI), ‘big data’ collection, and machine learning, according to expert Dr Brent Richards, Medical Director of Research Commercialisation and Adj. Professor of Critical Care Research at Gold Coast Hospital and Health Service.

Dr Richards has seen the enormous potential of MedTech innovation firsthand, including AI integration in hospitals, and is working on ways to positively change patient experiences, optimise clinicians’ priorities, and future-proof the healthcare system.

To drive his MedTech vision for Queensland forward, Dr Richards relies on an expert team of data scientists, clinicians, entrepreneurs, software developers, business experts, and industry influencers at Lumina, within the Gold Coast Health and Knowledge Precinct where he is based.

‘MedTech’ is devices, tools, equipment, and software that help diagnose and treat patients

MedTech refers to technology and innovation that is used for diagnosis, patient care, treatment, and general improvement of a person’s health. It can cover a broad range of innovation areas, including medical equipment, devices, machines, software, and tools that are common in a hospital setting, but may extend into medical innovation in the community as well.

Within the fast-moving world of MedTech, it is artificial intelligence (AI) technology, data, and computer science that are of most interest to Dr Richards, and the role they play in supporting clinicians, enhancing patients’ care, and optimising systems within the healthcare sector.

AI is transforming traditional methods of healthcare and how hospitals use their data

A practising Intensive Care Specialist, including a 15-year tenure as Director of ICU at the Gold Coast University Hospital, Dr Richards is also a co-founder of the IntelliHQ, a not-for-profit organisation. He established IntelliHQ to create and support an innovation hub for the health sector, that promoted education, research, investment, and commercialisation of technologies within AI and machine learning. IntelliHQ is headquartered within Cohort Innovation Space, a major Lumina tenant, and partners with Gold Coast Health, industry, and universities; all of whom are committed to transforming healthcare through AI and machine learning.

Artificial Intelligence, and its subfield of machine learning, in simplified terms, is about combining computer and data science to enable big data problem-solving. In healthcare settings like a hospital, AI and machine learning can be used to develop a system to read data (such as patient medical information) and use it to learn, through identifying patterns. The amount of data used to train these systems can be enormous – with some systems needing to be exposed to millions of examples to learn a specific task. However, once the system learns the tasks, the benefits can be huge, as the system can make connections and predictions much faster than a human brain. In a busy public hospital setting the potential is endless.

“AI innovation has major, positive implications on patient experiences. Historically we have been held back by slow methodologies in our health system. That’s been an issue. The answer is in valuing data, and collecting data, and then extracting information that empowers clinicians, patients, and the healthcare system” Dr Richards explained.

AI data systems in hospitals process real-time patient information that help doctors monitor and treat illness

One of Dr Richard’s most ambitious projects is in collaboration with IT company Datarwe, and the Intensive Care Unit (ICU) within the Gold Coast University Hospital. Together they have developed a data platform called the Clinical Data Nexus (CDN), which helps medical researchers, clinicians, and innovators to develop new diagnostic tools and treatment options for patients.

The CDN captures patient information from acute healthcare monitoring devices such as ventilators and heart rate monitors, as well as medical records, then uses advanced tools to process the data. The information is displayed in a virtual dashboard of real-time patient information in the ICU that helps clinicians monitor, treat, and plan resources for critically ill patients. The Dashboard is a key communication tool for hospital teams to visualise each patient’s clinical status and improvement or decline, as well as operational measures such as predicted discharge, bed availability, and staffing.

“To continue to provide optimal healthcare in the future, with a rapidly growing ageing population and finite budgets, we have to find other ways to solve problems in healthcare. This is where AI can come in and make a difference; it has so many potential applications across health.”

AI technology can connect individuals and communities to more accessible, personalised healthcare

Dr Richards often talks about the ‘democratisation of medicine’ to help explain how the community will benefit from AI integration in healthcare. The ‘democratisation’ implies more knowledge, convenience, and empowerment of patients to become more involved in their own care, and virtual access to specialist level knowledge where they are not currently or routinely available.

Dr Richards has been a part of multiple health innovation projects that use AI to empower patients. One recent initiative involved using medical imaging AI to diagnose ear infections in remote indigenous communities. Dr Richards and his collaborator Dr Kelvin Ross heard about a study outlining the high prevalence of ear disease and associated hearing loss among Aboriginal and Torres Strait Islander children. They predicted they could produce a solution using medical imaging AI.

“In some indigenous communities, there is a health team that comes in only once every three months. But ear infections don’t occur every three months. Nobody has ever been in a position where we have individuals examining their own ear. What it means is you’re taking the locus of opportunity and responsibility, right down to the patient level.”

They created an ear infection diagnostic prototype using an an-off-the-shelf otoscope and developed a catalogue of ear infection imagery to classify what was ‘healthy’ vs what was infected. A person could then investigate their own ear, using the otoscope, collect an image, and find out if they have an infection.

“The AI can direct them how to position the otoscope correctly to capture the right picture. Then they simply take a picture and receive the result via their phone, or they can choose to share it to the Cloud; for example, sharing with a clinician online.”

AI in medical image analysis and has optimised patient treatment across different diseases and conditions

Dr Richards also described a real-world example of how AI in medical imaging analysis can support neurologists to provide the best care for patients with epilepsy, using face-mapping technology.

“I was looking for complex partial seizures and face movements, that are difficult to see and not very well classified. I found a piece of camera software called Face Mesh, which sits on top of TensorFlow (an open-source platform for machine learning), that can map 468 different points on a human face, and then can measure their movement,” he said.

“It means you can take an analogue system, which would be me physically looking at a patient’s facial movements, and change it into a truly digital system. We are progressing from relying manually on humans and starting to think about how we can turn that human knowledge to collect the data, learn the data sets, do the AI and then start to move those AI-based algorithms to more healthcare settings.”

“We can use AI to take on some of this work, such as classifying imagery, and empower doctors to make more efficient use of their time and patient priorities.”

Successful integration of AI in healthcare requires a connection between different spheres of expertise

Solving complex health problems using AI integration is not easily achievable through one individual’s big idea – it requires teamwork and a combination of different skill sets.

“If you talk to the average clinician who is taking a clinical idea and moving it into data science, the two worlds are far apart. I think of it as different ‘windows’ of knowledge, that need to come together; and one of them has to expand its own window to do that. That is the role I play, in the middle, connecting the ‘windows’,” Dr Richards said.

Dr Richards thinks the ideal team should include experts such as software engineers, data scientists, and someone with clinical knowledge. He also emphasised that the team needs to be prepared to move at a higher risk level, and faster than they may be used to in the healthcare sector, which is historically highly risk-averse. This is why inviting commercial expertise into the team is also important. “You need to bring in the commercial side too, and think about more than just your community of patients. Think about the business community, and how involving them in the project can ultimately benefit the whole community and get solutions to patients faster.”

Training for health professionals is important too. IntelliHQ recently announced the launch of their national Big Data Capture, Management & Analysis, and Artificial Intelligence training program, to educate clinicians and medical researchers about utilising data and AI tools.

Lumina’s HealthTech accelerator program, access to mentors, networks, and facilities help drive MedTech forward

Except for a few heavyweights such as Cochlear, Australian MedTech companies are typically young and small. Innovation hubs such as Lumina are essential to help nurture their big ideas and link innovators with experienced, well-connected experts who can drive the idea forward.

Dr Richards spends a lot of time within Cohort, the start-up innovation space within Lumina, where he is a mentor to HealthTech start-ups participating in the LuminaX accelerator program. As part of his mentorship role, he can introduce promising start-ups to experts within the clinical and commercial sectors of his network, to help them access the various expertise needed to progress their ideas.

“As a start-up, it is likely not enough to rely on your own networks. Within Lumina, you are suddenly one or two contacts away from a beneficial connection who could partner with you, or discuss your idea, and understand where the holes in the market are, and where the competitive opportunities are. Otherwise, you would spend a lot of time (that you don’t have), doing it all alone, and finding your connections.”

Access to facilities and institutions, such as universities and clinical practices or hospitals, is also important in MedTech, as credible research, and evidence-based solutions are essential to ideas being accepted within the health sector. At Lumina, tenants work close to Griffith University, which includes the Institute for Glycomics and ADaPT (Advanced Design and Prototyping Technologies Institute), and the Gold Coast University Hospital. Tenants within Cohort Innovation Space can also take advantage of a purpose-built AI lab.

“AI is changing the locus of opportunity and responsibility in healthcare. To date, it has been institution-centric and provider-centric for those who work in the existing hospital system, but now we are progressing to being patient-centric, thanks to the power of AI,” Dr Richards said.

“The opportunities are endless.”