The hospital’s cardiology department hopes to roll out an AI tool to help rapidly diagnose heart attacks and speed up treatment.
The Mater Misericordiae University Hospital has today (17 June) launched the Centre for AI and Digital Health, to develop artificial intelligence (AI) solutions to improve and speed up patient care.
The new hub, which is based at the Pillar Centre for Transformative Healthcare, will oversee the adoption of AI tools in the hospital and coordinate collaborative AI research across the various departments in the hospital.
According to the Mater, AI solutions have already contributed to the reduction of emergency department waiting times and optimised hospital workflows.
The radiology department introduced AI in 2023 to quickly identify suspected conditions, such as stroke and fractures, and, according to the team, it has an accuracy rate of more than 90pc.
The cardiology department plans to roll out a new tool that supports early heart attack diagnosis.
Speaking to SiliconRepublic.com, consultant cardiologist Prof Joe Galvin says he thinks the greatest use of AI in cardiology is in helping with the interpretation of tests. AI can assess various tests, including electrocardiogram (ECG) and cardiac ultrasound, and see abnormalities that the human eye just doesn’t see, he says.
One important application of that ability is to help patients having a heart attack. Galvin explains that when a patient presents with chest pains, they receive an ECG.
In this test, the clinicians look for ST segment elevation, which tells them that the patient is having a heart attack due to a blocked artery. The team then aims to get the patient to the cath lab to open the blocked artery within 90 minutes to minimise the damage the blockage causes to the heart muscles.
The problem for medical teams is that ST segment elevation occurs only about 40-50pc of the time, so for the other 50-60pc of patients suffering from this type of heart attack, they do not get the same rapid treatment and instead must wait for further tests.
The AI tool that the Mater hopes to pilot in the coming months identifies the blocked artery in about 90pc of patients, doubling the number of patients that can receive rapid treatment.
Galvin says that early intervention for this type of heart attack is proven to reduce patient mortality rates, though it is too early to say whether this AI tool will also reduce mortality risk. “We don’t know yet for the AI ECG, we’re waiting to get that data, but I think it’s highly likely that it will reduce mortality,” Galvin says.
Another project the cardiac team is hoping to work on, Galvin says, is the use of AI to investigate potential causes of sudden arrhythmic death syndrome (SADS) to identify risks for the families of those who have suffered from this condition.
AI hub brings hospital teams together
Paul Banahan is senior AI research fellow at the Mater Hospital, Ireland’s the first dedicated AI researcher in a clinical setting.
In his four years at the Mater, Banahan has worked on a number of AI research projects, including on a generative AI tool to convert CT images into synthetic MRI images to enhance emergency spinal care.
He explains to SiliconRepublic.com that it would not be possible to conduct a lot of this research outside a hospital setting.
“Some of the work that I’ve done – the generative AI project for spinal injuries – couldn’t really have been done to the degree that it has been done outside of the hospital because it involved so many specialties and expertise from across the hospital.
“There’s a type of research we can do here that academic bodies on their own can’t do and can be quite time-consuming and expensive for [those bodies] to do.”
Banahan is currently completing a PhD at University College Dublin that looks at generative AI models for brain imaging, “with the aim of being able to forecast the progression of stroke and brain injuries” and understand how the AI models interpret the data they receive.
Ethics of AI in healthcare
AI use inevitably generates ethical and privacy concerns, particularly when working with sensitive healthcare data.
Banahan says that research ethics and GDPR are always top concerns and that when AI projects are undertaken there is a “rigorous process” to get approval from the hospital’s research ethics committee, which requires the project team to document in detail what they are going to do and how, and ensures they have the tools in place to do the project safely.
Josephine Ryan Leacy, CEO of the Mater Hospital, said in a press release that the new centre is focused on “ensuring that AI is developed and deployed in a way that prioritises patient safety, transparency and real clinical benefits”.
In terms of diagnostics, Galvin emphasises that AI is very much an additional tool to help clinicians. The doctor is not handing over controls to AI.
“I would see this very much as just being an additional tool that will contribute knowledge and that will help the doctor make up their mind as to what the problem is and how to proceed, but it’s still the clinician who makes the decision and has the responsibility.
“It’s just that they have another diagnostic aid, and one that that does appear to be exceedingly accurate.”
Don’t miss out on the knowledge you need to succeed. Sign up for the Daily Brief, Silicon Republic’s digest of need-to-know sci-tech news.
