Aengus begins by outlining the inspiring objectives of the startup medtech he co-founded with his brother Dimitry Tran, Harrison.ai. “We’re an Australian-based AI technology company & we are clinician led. The goal for Harrison.ai is to build cutting edge AI technology and putting that into the hands of clinicians with the explicit goals of increasing accuracy of diagnosis, improving time to diagnosis as well as access to care in places where it may not be available.”
“We want to do that by building these tools in a very quick turnaround time.” The team started by building chest x-ray and CT brain products for radiology (through venture, Annalise.ai). More recently moving into the field of pathology with Franklin.ai.
They then discuss the founding of Harrison.ai – ‘not your average startup’. Aengus explains, “My background is a medical doctor by training, I went to medical school at UNSW and during this time I became really inspired by the idea of being able to help more patients than you can physically see in your lifetime as a doctor.”
“I see AI as one of the enabling technologies that will allow us to scale the knowledge & capabilities of clinicians beyond what they can achieve in their lifetime.”
When asked about how radiologists across Australia are utilizing the technology, Aengus talks about a chest x-ray diagnosis tool, Annalise Enterprise CXR.
“We built a product called a chest x-ray solution. As a patient if you have a chest x-ray for whatever reason, you typically face an anxious wait of days to weeks before you receive the results from the chest x-ray analysis.”
He also highlights the potential for missed diagnosis and human error in the manual process as well, “The clinicians who are doing this interpretation could be reading this at night, after a long shift, or perhaps they could be distracted by a phone call.”
“So what we’ve done is to build a tool that can detect 124 different features of the chest x-ray using the power of AI. We can analyse this chest x-ray to provide a second opinion to the radiologist.The software enables the clinician to both improve their diagnostic accuracy but also prioritise their efforts – looking at the cases with the most important findings first.”
“It’s like being able to go through your email, identify the most important email and reply to that first. Same is true here. Radiologists using this AI can identify the most critical x-ray (or CT brain) and read that first.”
So what’s next for Harrison.ai?
“If we look globally, there’s an increasing shortage of clinicians.”
“Being a doctor by training myself, I know it takes a long time to train – 6 years in med school and another 6 years of specialty training, so it takes anywhere from 10-15 years to produce a specialist radiologist or pathologist. And we know as a society that we will not be able to keep up with the increasing demand for healthcare.”
“The role for AI is to scale the knowledge of healthcare and the accurate diagnosis but add the power of computing and the scalability of software. As society progresses, we will need this underlying technology to help us combat current and future health challenges. By doing so, that’s how we can improve the standard of care while maintaining a reasonable cost of care that most governments in the world are struggling to maintain.”
Subscribe below to keep up to date with the latest news & events from healthtech & Harrison.ai.
