How AI is Changing Patient Care: A Conversation with Professor Peter MacMahon
In the latest episode of Transforming Healthcare – The People Behind the Progress, we sit down with Professor Peter MacMahon, Consultant Radiologist at the Mater Misericordiae University Hospital and Clinical Professor at University College Dublin, to explore the real-world impact of artificial intelligence (AI) in clinical practice.
As a pioneer in emergency imaging and a key voice in the adoption of AI in radiology, Peter brings clarity and candour to the opportunities and limitations of AI in today’s healthcare landscape. His insights offer a grounded view of how digital tools are reshaping diagnostics, patient care, and the future of clinical decision-making.
Radiology: Leading the Digital Charge
Radiology has long been at the forefront of digital transformation. From the early shift to digital X-rays to today’s AI-powered triage tools, Peter notes that radiologists have been well-positioned to adopt new technologies thanks to years of working with digital imaging systems.
“Radiology is one of the first specialties to go fully digital. That gave us a head start in implementing AI, because AI needs data, lots of it and digital infrastructure to work.”
AI in Action: Faster, Smarter, Still Human
At the Mater, AI is already deployed to assist with fracture detection in X-rays. Rather than making diagnoses, AI highlights areas of concern serving as a “second pair of eyes” to support clinical staff. This is especially valuable in busy emergency departments or during night shifts, when rapid decision-making is crucial.
“AI won’t replace clinicians it helps them. It speeds up diagnosis, flags things we might miss, and allows us to prioritise urgent cases faster. That ultimately benefits the patient.”
Peter emphasises the importance of clinical oversight. AI tools are powerful, but not perfect. The real value lies in using them to complement human judgment—not replace it.
Connecting Care with Digital for Care
AI is not a standalone solution it must be part of an integrated, system-wide approach. Peter sees AI as a natural extension of the HSE’s Digital for Care programme, which aims to create safer, more connected, more effective care pathways.
“The real power of AI will come when it’s fully embedded in national digital systems. Imagine a tool that reads the scan and the medical record, and instantly routes a patient into the right care pathway without delays or duplication.”
This future isn’t far off. Pilot projects in stroke care are already demonstrating how AI, digital imaging, and cross-team collaboration can create fast, seamless, and life-saving patient journeys.
Guardrails and Governance
Peter is clear-eyed about the risks. He stresses the importance of training, particularly for junior staff who might over-rely on AI outputs. AI should assist decision-making, not replace clinical accountability.
“It’s about balance. Just like medications carry risk, so does AI. But we don’t ban treatments because of side effects, we manage them. The same approach applies here.”
Governance, standards, and continuous evaluation will be essential as AI tools expand across diagnostics, treatment planning, and service coordination.
Looking Ahead: A Human Future
Peter’s hope is that the next generation of clinicians will inherit a system where digital tools are embedded, evaluated, and impactful not chosen for novelty, but for patient benefit.
“We need to ask: is this tool clinically useful? Does it support the healthcare worker? Does it reduce burnout? Most importantly does it help the patient?”
He also offers reassurance to patients who may be wary of losing the human touch:
“AI should lead to more time with clinicians, not less. It’s there to support safer, faster, better care not to replace the people delivering it.”