Screenshot 2024-07-05 at 4.12.36 pm.png

Image generated with DALL-E 3

"As AI continues augmenting the expertise of healthcare professionals, look for it to go further and do nothing less than enhance humanity," – Fei-Fei Li (2024), Computer Scientist from Stanford School of Engineering.

Discipline-specific considerations of generative AI

Ethics and privacy (data security): AI is useful for learning and assisting in case scenarios, but it is crucial to protect sensitive data and ensure patient confidentiality. Health students have a responsibility to ethically manage data, not to provide AI tools with highly protected data such as personally identifiable medical information, especially in real clinical scenarios and placements - to protect their own, and others’ information rights.
Bias: Learning in specific disciplines is heavily determined by the needs and resources of the schools within the faculty of Medicine and Health. AI systems are limited by the data it’s trained on. Therefore, students should critically evaluate the biases in AI tools, and use it as a supplementary tool to the University’s resources.
Reliability: AI tools lack the specific training on the detailed requirements of medicine and health units, and therefore not accurately interpreting the nuances of specialised medical fields. Students should approach AI as a tool that provides broad insights, rather than a source for definitive answers.
People orientation: In a field that emphasises individualised care and human interaction, prioritising patient needs and ethical considerations is essential. AI cannot replace the necessary people skills in healthcare.

Real-world examples of generative AI in the discipline

Diagram illustrating AI and robotics' role as part of the healthcare eco-system.
(Image source: 'No longer science fiction, AI and robotics are transforming healthcare,' PwC)

Harvard Medical School implemented the model 'SISH' (self-supervised image search for histology) which acts as a search engine for large databases of pathology images, having the potential to identify diseases and determine patient therapies.
AtomNet uses structure-based drug design to predict the bioactivity of small molecules for drug discovery applications.
IBM's WatsonX Assistant is a virtual assistant which has 'life-like' conversations with patients, catering to their needs to provide personalised and precise responses in healthcare contexts.

Some prompts to help in the discipline

Connecting new concepts with previous knowledge
Preparing questions, then practicing them by Generating multiple choice questions, and Practicing through cloze passages.
Breaking down topics using the Feynman Technique to consolidate understanding
Managing and planning your time by Planning your study and creating a Study Plan and Timetabling for Exam Preparation
Training with simulated scenarios to practice your responses and receive instant feedback
Overviewing a new concept to assist in your learning

Ensure you check out all the different ideas of how generative AI can support your learning.