The UK healthcare sector is undoubtedly combining practices with the use of AI to make technological advances within the NHS.
In the past year, the Government has awarded £13.7 million to the development of AI and Robotics to “capture… the value in algorithms generated using NHS data”. The use of AI however, has been met with criticism by some as a research has proved that hackers can use the artificial intelligence to actually attack medical scans, thus making the NHS’ technological revolution exposed to new threats.
Views on AI in healthcare:
In October 2019, Matt Hancock (Health and Social Care Secretary) renewed the focus on the use of AI with the “Future of Healthcare” report which will enable access to real-time data and set standards for security and interoperability. In June 2019, Simon Stephens, Chief Executive of the NHS said that AI will be a “huge part of life”, so he wants to “kick off a global call for evidence”, however the biggest challenge of using AI in healthcare is trust, as doctors don’t know if the AI is doing the correct job, and patients don’t know how reliable it is.
Recently, Parliamentary Science and Technology Committee Chair Norman Lamb stated: “Infrastructure for collecting sharing and accessing data needs to be improved. Resolving the ethical questions surrounding AI in healthcare settings will be crucial, including the setting the right regulatory framework.” This clearly shows that the NHS needs to make sure that the data fed to the AI algorithm is wholly accurate to minimise any chance of mistakes.
But why use AI in healthcare?
AI can not only be used to reach ground-breaking discoveries and increase the effectiveness of patient care, it can also make the NHS’ administrative functions much more efficient and thus will deal with routine administration, saving money which can then be funded elsewhere. The BMA has found that doctors spend 15% of their time on administrative work, however this can be streamlined massively by using AI thus time spent on bureaucratic work will be shifted to extra time to care for patients.
However challenges AI faces is that some technology can produce new data which is why the GDPR came around which has transparency and accountability at its core to make sure patients are safeguarded. The healthcare sector is particularly vulnerable and therefore the use of AI must be welcomed with slight caution due to unforeseen situations where data can be automatically made by AI. The size and availability for sensitive data records must be safeguarded as it is a rich source of valuable data for hackers. There has been a recent partnership between Amazon Alexa and the NHS which has split opinion among AI experts and data ethicists as Tech Expert Mathana Stender explains: “the sensitive data holding of the NHS are a form of critical social infrastructure” but this was handed to a foreign company showing that regulation and transparency is vital if AI was to be combined into the NHS. Since the scandal of DeepMind in 2017 (where there was an unlawful transfer of 1.6 million patient records to Google’s AI system), there has been a decrease in public acceptance towards the use of AI, so the issue of trust needs to be addressed.
Furthermore, implementing AI into healthcare will be challenging as medical data is extremely sensitive and private. Therefore, regulation will also need to be set in place regarding algorithms however, the medical professional will also need to be trained to explain how and why AI is being used, so they can reassure the patients. This also requires trust from the doctor to believe that the AI can accurately and correctly amplify their medical knowledge. In addition to being clinically effective, AI must also be cost-effective to truly benefit the healthcare sphere.
The Government has taken steps to make sure that AI use is tracked and safe, by issuing the NHS Code of Conduct (Feb 2019) which ensures that only the best and safest data-driven technologies are used by the NHS which will protect patient data. The code will also encourage tech companies to meet a set of standards and promote the UK as the best place to invest in health-tech. The House of Lords Committee on AI has recently identified “transparency” and “explainability” as key requirements if AI is to become an integral tool in healthcare and further, society.
AI is already being used in healthcare, with an example being the Moorfields/Deepmind partnership where one million anonymised eye scans were shared with Deepmind under a research agreement that began in mid-2016. This collaboration helped quickly detect eye conditions and prioritise patients in order of urgency. Results also matched the accuracy of expert doctors with over 20 years of experience, thus showing clear benefits. Other AI benefits can already be seen through Oli the Chatbot in Alder Hey Children’s hospital that has provided improved levels of care as it allows the children to ask the Chatbot questions that they may feel awkward asking otherwise. Furthermore, evidence shows that a high proportion of mammograms yield false positive results when interpreted by radiology leading to one in two healthy women being told they may have cancer, but with the use of AI, mammograms can be interpreted 30 times faster than humans and with greater accuracy thus reducing the need for unnecessary biopsies and the concern of a misdiagnosis.
The NHS has a clear advantage when it comes to using AI as it is the biggest healthcare organisation in the world with over 1.7 million staff and 1 million patients seen every 36 hours, thus can provide huge amounts of data to improve accuracy of AI algorithms.
The UK ranks fourth in creating the right conditions for home-grown AI to flourish, however different areas must work together to share and integrate findings and innovation in order to provide better services.
The future for AI in healthcare must definitely be considered as a strength to innovate, however will certainly be met with criticism and worry regarding trust thus regulations will almost definitely be put into place.