AI and medicine
underway in Northern Lincolnshire and Goole. 7 The NHS is also exploring the Virtual Wards programme, where patients are provided with, and trained how to use, portable healthcare equipment such as pulse-oximeters and blood pressure monitors. They are able to directly relay that information to staff at their local hospital, in the hope that interventions can be made earlier, before a patient needs admitting to hospital. This is currently being used for pre- and post-operative patients to assess their recovery. 8 The NHS is revolutionizing the way outpatient and pre- and post-op recovery works by putting the patient in charge of their recovery and enabling them to better look after themselves. All with the added benefits of saving bedspace and reducing workload for NHS staff. There are issues with AI. The first is the huge cost involved and the time it takes to develop safe, secure and working systems. The use of AI also raises ethical issues: for instance, how will informed consent concerning how these machines and algorithms work? How is data used, both from the patient and also from the clinician’s side ? One of the most challenging predicaments is the use of so called ‘Black Box’ programmes, in which humans are unable to understand how an algorithm reaches its conclusion. These scenarios usually ‘ result from algorithms that utilise non-interpretable learning techniques that are difficult for clinicians to understand fully ’ . 9 This raises numerous ethical concerns from the perspective of a patient, as they have the right to ask how a specific drug or programme works when it is making what could be a serious diagnosis. Furthermore, doctors might not totally understand how the tool works due to the complexity of these algorithms, which could lead to problems with confidence in medical professionals and may undermine the relationship between the doctor and patient. Additionally, under the EU’s GDPR (General Data Protection Regulati on), a patient has the right to refuse disclosure of their personal data, which could mean that they are unwilling to be diagnosed via a process using AI. It is important that a balance struck in order to achieve the optimum level between the effective and accurate use of machine learning and AI programs and the privacy afforded to patients, so that the doctor/patient relationship isn’t damaged. Artificial intelligence could have a phenomenal effect on patient care, most notably in diagnostics, where efficiency will be dramatically improved. This will have numerous benefits not only for the NHS, but for healthcare providers around the world, in reducing workload, particularly in a post-COVID world where these optimizations are needed more than ever. However, the ethical and legal dilemmas raised show there are hurdles that need to be examined before this revolutionary technology becomes the norm. The human emotions and empathy provided by the care of a doctor or nurse can never be fully replicated by a machine, but AI offers a fantastic tool to healthcare providers in the future.
Bibliography
Bora, P. 27 Companies changing health outcomes through AI. https://www.digitalauthority.me/resources/artificial- intelligence-in-healthcare/ Consulted 13/7/22 Chadwick, J. DeepMind’s AI spots early signs of eye disease. https://www.zdnet.com/article/deepminds-ai-spots-early- signs-of-eye-disease/ Consulted 13/7/22
7 See https://www.england.nhs.uk/2022/08/nhs-trials-smart-goggles-to-give-nurses-more-time-with-patients/ 8 See https://www.england.nhs.uk/virtual- wards/#:~:text=The%20NHS%20is%20increasingly%20introducing,devices%20such%20as%20pulse%20oxim eters 9 See Gerke, S. et al https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7332220/
187
Made with FlippingBook - Online catalogs