Artificial intelligence (AI) mimics human behaviour and cognitive functions and is currently being used to bring substantial changes to the NHS. The use of AI in the health services is on the rise, the idea is to model a real living organism with a brain; a sensory system which collects data of every aspect of the healthcare service, helping to care for families and freeing up time for health professionals.
As funding declines it is unlikely to meet increasing demand, bridging this gap will require creative thinking around the use of AI. Since the NHS has formed people are living longer and chronic diseases now account for 70 percent of the healthcare spend. Due to issues around the level of knowledge and complexity, healthcare professionals can be supported by AI to ensure that they can meet the ever-increasing demands. This is where AI shows its strength, as it can support professionals whilst cutting costs at the same time.
Health Secretary Matt Hancock said: “Artificial intelligence will play a crucial role in the future of the NHS and we need to embrace it by introducing systems which can speed up diagnoses, improve patient outcomes, make every pound go further and give clinicians more time with their patients”(Telegraph, 2018).
‘AI has now been shown to be as effective as humans in the diagnosis of various medical conditions, and in some cases, more effective.’ (Loh, 2018)
General Practice is the subject of change
Due to the changing nature of general practice surgeries, CCG’s in London gave rolled out a trial of a new AI Ap, with easy access to large datasets, access patient records and to retrieve a diagnosis in a matter of second. GP’s are also able to share information between one another in the hopes to speed up the diagnosis on each patient. This can deliver better outcomes through better prediction, detection and management of health conditions. Applications such as these can improve the quality of care and patient outcomes whilst more importantly, reducing the NHS costs.
Every year, 2.5 million scientific papers are written in medical journals, which is impossible for any Clinician to keep up-to-date with. IBM’s Watson can be used to crunch huge amounts of data. More specifically, it can read and process existing medical articles along side patient data to aid in the diagnosis of patients.
Patients are also embracing new technologies and increasingly expect their care to be supported by it. Most people say they would use virtual GP technology over going to the Doctors, to diagnose minor ailments and on-going long-term conditions. (Castle-Clarke, Sophie 2018)
In polling 2,000 UK adults, (aged 15 and over):
- 63% of respondents said they would use virtual assistant to help diagnose small ailments.
- 55% said they would consult with their GP over virtual assistant to help with on-going problems.
- 54% said they would consult with their GP using virtual assistant when it’s an emergency.
What will new technology mean for NHS and its patients. (Castle-Clarke, Sophie 2018)
Will robots diagnose and be used more frequently?
Currently AI has been designed to process blood sugar levels for those with diabetes, it learns about the individual then sends guidance data to help them manage their disease. Patients can look after their own care as well as their healthcare team being able to access and monitoring data.
AI would require access to all these types of data across the NHS to ensure that it is informed enough to provide accurate diagnosis. Although this has the potential to significantly transform the NHS, currently, the data collected is not shared at a national level and therefore cannot be linked together to create large datasets for AI to access.
We have already seen the introduction of AI, however, until the issue with data has been resolved, this can only potentially be used as a locality-based option.
Studies conducted using AI – The results
AI has shown to be effective in finding diagnosis in various medical conditions. Neuroscience has benefitted from AI. A deep-learning learning algorithm used MRI of the brain of individuals 6 to 12 months old to predict the diagnosis of Autism. The results were staggering, the AI was able to positively predict at 81%. In another study a machine learning algorithm managed to obtain an accuracy of 84% progression of those with progressive dementia. (Hazlett, et al, 2017)
In ophthalmology, an AI-based grading system was used to screen fundus photographs obtain from those with diabetes. The AI identified, with high reliability, (94% and 98% sensitivity and specificity), to determine cases that should be referred further to an ophthalmologist for further treatment and evaluation. (Gageya, et al, 2017)
Will AI completely take over the role of GP Doctors?
Currently AI is not able to replace GP’s, tests have shown that humans can more accurately diagnose a patient. However, when it comes to treatment planning, AI has a significant advantage from a time saving perspective. (Loh, 2018)