Jack Breen’s talk about the use of Artificial Intelligence for the diagnosis and treatment of Cancer was educational and very relevant to modern society due to technology’s rapid evolution in recent years. Jack has an undergraduate and master’s degree in maths, as well as an accompanying master’s degree in medical AI from the universities of Nottingham and Leeds making him an esteemed expert in this area, enriching us with his knowledge on AI and further medical research.
Jack went into detail regarding cancer diagnosis looking at how it relies upon physical examinations, radiology, pathology and blood testing to assess the internal functioning of the body. Then on to state how AI could provide more depth about the relationships of symptoms and ultimate justification regarding the causes of the symptoms, becoming more accurate than the existing risk assessment tables used by GPs. The benefits of AI were then discussed by looking at the workload of diagnosing cancer patients and how creating a treatment plan is overbearing for pathologists so AI would provide a fast and accurate second opinion about the cancer diagnosis.
Ultimately concluding that whilst AI is a revolutionary advancement in cancer treatment, the application of these technologies within the large scale of the NHS is impractical and expensive – in addition the constant risk of online safety to retain patient confidentiality is significant and hard to control.
Q&As:
What would the alterations in speech be as a result of cancer?
Typical voice changes are a deeper voice, a change in voice patterns like pitch and small rhythms in language that a human would not necessarily pick up on. Even an Alexa, which listens to your voice daily, would have the ability to recognise when there is a worrying change in your voice.
Would there be any ethical dilemmas between humans and the AI?
There is the issue of who to blame if AI fails and experts could be taken to court if they override the AI advice and fail to pick up on cancer within an individual. However, radiologists have the highest form of knowledge and experience in this field so would have the ultimate decision over the AI.
In regards to building the AI models, how is the trial conducted to allow the most accurate knowledge being inserted into the AI system?
When conducting the trial, the opinions from the best experts in the country/world would be sourced as they are highly experienced. In addition, we would seek out specialised experts who have a deep understanding about their particular field (e.g. gynaecology) and use multiple of these opinions to make a joint decision.
Ms. Wingrove, Teacher of Biology
