Artificial intelligence could revolutionise healthcare by allowing doctors to spend much more time with their patients, according to a leading health and technology expert at Microsoft.
Dr Kenji Takeda, Director of Academic Health and AI Partnerships at Microsoft Research in Cambridge, told the All Party Parliamentary Group (APPG) on Heart and Circulatory Diseases that helping the medical sector is one of the most important uses of AI.
A survey by the group found that people in the UK support using more technology in the healthcare sector, with 85% of respondents backing the use of AI in diagnostics and treatment, and 86% saying they were happy for their anonymised health data to be shared to better diagnose medical conditions.
Takeda was joined at an APPG event at the Houses of Parliament recently by Secretary of State for Health and Social Care Matt Hancock and Simon Gillespie, Chief Executive of the British Heart Foundation, among others. A report they released revealed that there are seven million people living with heart and circulatory diseases, such as coronary heart disease and vascular dementia, in the UK and they cause a quarter of all deaths. It found that there was huge potential for AI to transform the lives of those people and a greater need for them to be included in discussions about the development and adoption of new technology.
Takeda said one way technology could help was by allowing doctors to spend more time with their patients.
“For every hour spent with patients, physicians may spend two additional hours documenting clinical interactions,” he said. “At Microsoft, we’re committed to empowering people on the frontline of healthcare to help them deliver the best care they can,” he said. “But we also want to empower patients, too because all of us touch the healthcare system at different points – a local pharmacy, our GP surgery, a hospital, and at home.
“We need to make sure that we’re making the best use of technology … and giving clinicians the time to really focus on patients. It’s what people deserve. Technology has to help deliver that.”
Takeda, who is also a Visiting Industry Fellow at the Alan Turing Institute and Visiting Fellow at the University of Southampton, highlighted a Microsoft project called EmpowerMD. This ambient clinical intelligence system captures conversations between doctors and patients, and integrates this with a patient’s electronic health records by synthesizing relevant information into the correct sections of a medical note. This allows them to spend more time talking to patients and understanding their needs, and less time updating medical notes.
“To truly revolutionise healthcare, data that’s currently locked away in single departments, systems, surgeries and hospitals, needs to be shared among medical professionals,” he added. “This requires data interoperability, and decision-makers in the NHS to adopt cloud computing as the secure digital platform so staff and researchers can act on information in real-time.”
An example of this is the Azure API for FHIR (Fast Healthcare Interoperability Resources) preview already being used on projects supported by Health Data Research UK for optimising at Optimising clinical data for research at Ormond Street Hospital, and the cloud-based integration project for rare disease research led by NIHR BioResource.
“The power of AI and machine learning can really make a difference in the UK by putting people at the centre,” Takeda said. “To achieve this, we need to design AI to be trustworthy and create solutions that reflect ethical principles that are deeply rooted in important and timeless values: fairness, reliability, privacy and inclusivity. Transparency and accountability are also key to building trust in AI in healthcare, and across all applications and industries.”
The APPG event, which also heard from Dr Indra Joshi, Clinical lead for NHS England’s Empower the Person portfolio, Hilary Newiss, Chair of National Voices, and Henry Smith MP, reinforced six key recommendations from its report:
- NHSX should set up discussions with charities and the public, to explore patients’ views and concerns about the use of AI in healthcare.
- Understanding Patient Data (UPD) should work with charities, patients and the healthcare sector to develop tools and resources for engaging the public on AI.
- Academic Health Science Networks should facilitate the exchange of information around new developments in AI between patients charities, and industry partners.
- NHS England and NHS Digital should explore the impact of wearables and AI on health inequalities.
- NHSX should work with UPD, charities, and patient organisations to ensure that policy development in AI is designed with the explicit purpose of understanding, promoting and protecting public values and that this is clearly and openly communicated.
- NHS England and NICE should develop standards for publication for AI research, providing trustworthy guidelines for researchers, the media and the public.
“While the focus of this inquiry is on heart and circulatory diseases, the recommendations and discussion are applicable more broadly,” the report added.
Hancock, who became Health Secretary in July last year after serving as Secretary of State for Digital, Culture, Media and Sport, agreed that technology could have positive implications across the medical sector.
“There is no doubt the technology that has been developed, including AI, has huge potential for saving and improving people’s lives,” he said at the event, pointing out that seven out of 10 people now survive a heart attack.
“Data improves the technology, it improves prediction and prevention and knowing which treatments work better. Using patient data improves treatment for everybody because every time there is a new data point, we can learn from that and improve.
“But to get the most out of this opportunity, you need people’s consent, and that requires trust. In order to have trust, we need to have strong rules around privacy and cyber security, so people know the data will be safe. We also need to explain why it’s important for people to allow that information to be used for research purposes.”
