RCSI was delighted to welcome a significant number of healthcare professionals and educators in attendance at its recent AI2MED webinar Intelligent Care: Bridging AI Innovation and Healthcare Education, presented by Dr John Sheehan. Dr Sheehan, a consultant clinical radiologist and specialist in digital health innovation, currently is co-chair of the American Irish Medical Summit and co-chairs the Digital Health Symposium at the United Nations General Assembly.
Dr Sheehan delivered an insightful exploration of artificial intelligence’s transformative potential in healthcare, vividly illustrating its applications across diagnosis, personalised treatment, and patient care. This was balanced by a candid assessment of some the critical challenges facing AI integration in medical practice and need for widespread education of clinicians and engagement with the wider public.
Dr Sheehan discussed the current state of AI in medicine, focusing on its ability to recognise complex patterns in multimodal biomedical data, which can aid in diagnosis, personalised treatment, and improving patient outcomes. Radiology stands at the cutting edge of AI adoption, with 80% of FDA-approved AI products concentrated in this domain. Dr Sheehan highlighted impressive examples of AI’s diagnostic prowess:
- Stanford’s ChestNet can detect pneumonia with accuracy surpassing board-certified radiologists
- DeepMind’s OCT AI identifies up to 50 eye conditions with a remarkable 94% referral accuracy
- Systems like IDX-DR enable early diabetic retinopathy detection without specialist intervention
Challenges and Ethical Considerations
While the technological promise is immense, Dr Sheehan was forthright about the significant barriers to AI integration. Regulatory uncertainty, clinical scepticism, lack of transparency, algorithmic bias, and data privacy concerns loom large. The EU AI Act’s classification of medical AI as high-risk underscores the need for transparent algorithms, rigorous human oversight, robust ethical frameworks and collaborative institutional approaches.
Generative AI and Quantum Computing
Dr Sheehan explored the potential of generative AI, such as ChatGPT, in healthcare, highlighting its applications in clinical documentation, medical imaging, and drug discovery. He mentioned Google’s Med-PaLM 2, which has achieved expert-level scores on medical licensing exam questions, and Epic’s partnership with Microsoft to embed GPT-4 models into electronic health record (EHR) systems for drafting patient messages and visit summaries. Generative AI can also produce synthetic medical images to augment training datasets and propose novel molecular structures for drug discovery.
Quantum computing, leveraging quantum mechanics, emerged as another frontier with the potential to transform drug discovery, personalised treatment, and healthcare logistics. Dr Sheehan discussed IBM’s collaboration with Cleveland Clinic to implement the first quantum computer system in a hospital group, and the applications of quantum computing in simulating complex biomolecular interactions, optimising clinical trial designs, and solving healthcare logistics challenges.
Educational Innovations and Recommendations
The webinar emphasised the need for digital and data literacy among healthcare professionals. Key recommendations included integrating AI into medical curricula, continuing education in areas such as AI interpretation, workflow integration, ethics, and data governance, and specialised programs to equip healthcare professionals with the necessary skills. He noted that institutions such as RCSI and Harvard are adding health informatics and digital medicine modules, while the NHS has launched the Topol Fellowship to train clinicians in digital health and AI.
He also highlighted the importance of engaging the public to demystify AI and address fears. He pointed to the opportunity afforded by RCSI’s new public engagement space, which aims to educate citizens about healthcare. Overall, the webinar underscored the need for collaboration, ethical implementation, and continuous education to harness AI’s potential in healthcare.
