Microsoft UK published its latest artificial intelligence (AI) report entitled “Accelerating competitive advantage with AI: How organizations are moving from experimentation to business impact” yesterday (1 October). The report can be regarded as the continuation of its counterpart published in 2018; both urge leaders of different business sectors who wish to stay competitive in the world stage, to adopt bold strategies embracing AI. As Cindy Rose, Chief Executive Officer, Microsoft UK remarked in the Forward, “if the message of our 2018 report, ‘Maximizing the AI Opportunity’ was ‘get started’, 12 months later it is ‘get serious’”.  

What does it say about healthcare? 

As the report was prepared by Microsoft UK, Britain and its National Health Service (NHS) as well as the surrounding European landscape were predominantly the focus. Overall, the report gives a positive tone in the use of AI in medicine and healthcare. As Stephen Docherty, Industry Executive – Health, Microsoft UK said, “we live in exciting times where the speed of technology adoption is rapidly increasing and there are multiple opportunities to use AI to benefit healthcare. Above all, we need to give clinicians back the gift of time while using AI to determine insight from the data we have”. 

According to the report, presently one in five AI start-ups in the EU operates in the healthcare industry, with a third originating in the UK. This is likely to continue in the future because of the £50 million in government funding allocated this year to support five new university-based healthcare AI centers in London, Glasgow, Oxford, Leeds, and Coventry. The report also revealed 46% of the healthcare leaders said their organization is now using AI. Even though it’s below the national average of 56%, there has been an encouraging rise of 8% as compared to a year ago. 

Of various domains, research level AI, robot process automation (RPA) and general automation, and voice recognition and touchscreen technology showed the biggest leaps of increase by 13%, 10%, and 9% respectively. It appears that the recognition of AI’s transformative potential for healthcare organizations is becoming widespread. The trend is not just reflected in the so-called “back office” processes but there is a genuine will to enhance patient care quality via the use of new technologies. “AI in healthcare is an extremely exciting prospect. It’s not about replacing staff, but allowing them to maximize their skills, be more efficient, spend more time with patients and, ultimately, get better outcomes”, added Darren Atkins, Chief Technology Officer, East Suffolk & North Essex NHS Foundation Trust. 

Research level AI, robot process automation (RPA) and general automation, and voice recognition and touchscreen technology showed the biggest leaps of increase.

What are the challenges? 

On the other hand, the report critically pointed out the use of AI in healthcare remains, largely restricted to smaller, localized pilot projects geared towards specific, practical outcomes. This reflects the industry is more interested in exploring the technology rather than embedding it into the system at scale. As such, bridging the gap between experimentation and implementation will be the next big challenge to address. 

This is going to be tricky especially healthcare places priorities on improving patient experiences and recovery and increasing efficiency while not compromising the welfare of its staff. The report found 96% of healthcare employees have never consulted their superiors about the introduction of AI in their organizations. 69% of them are still undergoing training to better their understanding of how to use the technology in their job. Shall the industry truly wishes to embrace AI at a great scale, these two areas ought to improve dramatically. 

When it comes to identifying bias and knowing what steps to take to address it, healthcare leaders are clearly lagging behind the national average

On top of which, over 37% of healthcare leaders expressed preparing usable data as their biggest challenge in scaling AI. With that, Laura Robinson, Senior Director for Healthcare, Microsoft UK explained, “the biggest flaw in the UK health system is that it has the data that a lot of companies would happily pay for but there is no way to guarantee clarity of ownership or governance”. Indeed, the industry know that information is available but in order to use them, there is a need for an effective and responsible plan. “The requirement of AI projects are data quality, engagement, integration, all things that are going to make positive changes,” Chris Carlin, Consultant Physician, NHS Greater Glasgow and Clyde added. 

Naturally, data usage leads to ethics. Patient privacy, cybersecurity, harnessing diversity and inclusion, and minimizing biases are key considerations that healthcare organizations ought to look out for while experimenting, validating, and deploying AI. Again, the report reflected a need to work towards these key areas especially when it comes to identifying bias and knowing what steps to take to address it, healthcare leaders are clearly lagging behind the national average. Overall, establishing a clear ethical framework, training industrial leaders and staff in responsible AI, highlighting any unethical practices and underlining the steps to take when problems arise, are critical shall the healthcare industry wishes to become truly AI-enabled. This is not only applicable to the organizations but for the UK as a whole. 

“If AI is going to work in healthcare, the industry needs to start with ethics. In fact, it’s so important that if you ignore ethics and education in order to speed things up, it will end up costing you. In the end – it may even lead to projects failing,” Valentin Tablan, Senior Vice President for AI, leso Digital Health (a cognitive behavioral therapy platform) said. 

Author Bio
synthetic gene empathy chinese artificial intelligence data medicine healthcare ai

Hazel Tang

A science writer with data background and an interest in the current affair, culture, and arts; a no-med from an (almost) all-med family. Follow on Twitter.