AI in healthcare – University of Glasgow perspective
Defining: AI in Healthcare
AI in healthcare refers to the use of computer systems to analyse complex medical data and support clinical decision-making. Although still a relatively new addition to clinical practice, AI is already being applied across a wide range of domains including diagnostics, treatment planning, drug discovery, and patient monitoring. Its potential is particularly notable in imaging-heavy fields.
In the next section, we will take a non-exhaustive look at some examples of AI in healthcare, specifically focussing on Scotland and the University of Glasgow. This is not intended to be a thorough overview of AI in healthcare, rather a bespoke Glasgow-centred introduction.
Clinical Applications in Scotland and the University of Glasgow
1. Early Detection of Lung Cancer in NHS Greater Glasgow and Clyde (RADICAL)
The RADICAL (Radiograph Accelerated Detection and Identification of Cancer in the Lung) study by the Digital Health Validation Lab is in collaboration with NHS Greater Glasgow and Clyde, the University of Glasgow, and Qure.ai. It assesses the integration of AI software into chest X-ray analysis to flag abnormalities suggestive of lung cancer, aiming to expedite further diagnostic procedures.
Image from RADICAL, Digital Health Validation Lab, 2025
2. Prioritising Head CT Scans in Emergency Departments (ACCEPT Study)
The ACCEPT (Assessing Clinical Effectiveness of Prioritising CT Heads) study by the Digital Health Validation Lab, evaluates the use of AI to prioritise non-contrast head CT scans in emergency departments. The goal is to determine whether AI can reduce report turnaround times and assist clinicians in managing patients with head injuries more effectively.
Image from ACCEPT, Digital Health Validation Lab, 2025
3. AI in Histopathology at the University of Glasgow
Researchers at the University of Glasgow and New York University have developed a pathology identification platform that employs Large Language Models to assist in the triage of cancer histopathology. This system, described as “learning the language of cancer,” has demonstrated the capability to identify disease signs in biological samples.
Image from “Mapping the landscape of histomorphological cancer phenotypes using self-supervised learning on unannotated pathology slides“, by Quiros et al., Nature Communications, 15, 2024.
4. Statistical Emulation in Cardiac Mechanics at the University of Glasgow
The University of Glasgow’s Bayesian Modelling and Inference section within the School of Mathematics and Statistics is using machine-learning tools (recall, the foundation to modern AI discussed in Introduction to AI, Week 1) to research cardiac mechanics.
Image from School of Mathematics & Statistics, University of Glasgow
TASK: Go on a research mission starting with the Wikipedia page on AI in Healthcare. This is a good starting point.
Below is the “see also section” from this page. Use this as a jumping-off point to explore the topic of AI in Healthcare more broadly and find some primary sources of interest.
- AI alignment
- Artificial intelligence in mental health
- Artificial intelligence
- Glossary of artificial intelligence
- Full body scanner
- BlueDot
- Clinical decision support system
- Computer-aided diagnosis
- Computer-aided simple triage
- Google DeepMind
- IBM Watson Health
- Medical image computing
- Speech recognition software in healthcare
- The MICCAI Society
- Algorithmic bias
Try to arrive at a research question you could explore. For example:
- “How can the effectiveness of a clinical decision support system be evaluated?”
- “In the context of healthcare, how does algorithmic bias differ from bias present in the data?”
- “What biases should we be aware of when implementing computer-aided healthcare triage?”
