AI could "reinforce health inequities" if data not diverse
Researchers have warned that new AI technologies could reinforce healthcare inequalities if the datasets they are based on are not representative of diverse communities.
Experts from INSIGHT, the UK's Health Data Research Hub for eye health, and University Hospitals Birmingham NHS Foundation Trust are describing this as 'health data poverty'.
In recent years, many academic and commercial organisations have been developing artificial intelligence and other digital health technologies based on publicly available datasets. But there is little information about how many datasets actually exist or the diversity of people and health conditions represented within them, which could lead to the development of technologies and products that only work for certain groups or countries.
The research team included experts from Moorfields Eye Hospital NHS Foundation Trust, the London School of Hygiene and Tropical Medicine, McGill University in Montreal, and Health Data Research UK.
Focusing on eye health, a leading area for digital innovation in healthcare, consultant eye specialist and director of INSIGHT Professor Alastair Denniston and his colleagues carried out a global search to explore the availability of publicly available datasets, and the extent to which they represented the diversity and needs of the world’s population.
They analysed 94 datasets containing 507,724 clinical images and 125 videos of eyes gathered from at least 122,364 people. They then created a comprehensive catalogue detailing the source of each dataset, its accessibility, and the populations, diseases and types of images represented within it.
The team found that most images came from populations in Asia and Europe, with very few datasets from large parts of the world such as sub-Saharan Africa (just one) and South America (two datasets). Looking closer, they discovered that information about the people within each dataset was generally poor, with basic demographic information such as age, sex and ethnicity being missing in more than one in five (20%) datasets.
This lack of geographical diversity and demographic data is a concern because technologies developed using data from one population may not work effectively when applied to a different group of people or part of the world.
There were also significant disparities in the types of diseases covered by the datasets. The majority of images were relevant to diseases such as diabetic retinopathy, glaucoma and age-related macular degeneration, mainly because these images are routinely collected as part of healthcare and screening in countries with advanced modern health infrastructure.
However, cataracts, trachoma and refractive error – which have been labelled priority eye diseases by the World Health Organization, and account for half of the world's blindness – were significantly under-represented. These conditions are common in low and middle income countries where digital technology could make a big difference making healthcare more accessible.
These conditions have not traditionally included imaging as part of their standard assessment, but there is increasing evidence that AI technologies could play a role. However the lack of relevant data for developing and training AI-based tools makes it less likely that researchers and companies will be able to develop products that could help.
The researchers are saying this situation risks creating ‘health data poverty’ for populations and countries that are under-represented in current health datasets.
“We hope that our catalogue will raise awareness of more diverse datasets for the development of AI-based health technologies" Professor Denniston said. "We need to act now to encourage health systems and researchers to invest in publicly available datasets to support research and innovation in areas that are currently data poor. Otherwise, we risk perpetuating a growing digital divide where healthcare technologies are only developed to benefit diseases, populations and countries with advanced data infrastructure.”
Check Point: Securing the future of enterprise IT
Cybersecurity solutions provider Check Point was founded in 1993 with a mission to secure ‘everything,’ and that includes the cloud. Conscious that nothing remains static in the digital world, the company prides itself on an ability to integrate new technology with its solutions. Across almost three decades in operation, Check Point, with its team of over 3,500 experts, has become adept at protecting networks, endpoints, mobile, IoT, and cloud.
“The pandemic has been somewhat of an accelerator in the evolution of cyber risk,” explains Erez Yarkoni, Global VP for Cloud Business. “We had remote workers and cloud adoption a long time beforehand, but now the volume and surface area is far greater.” Formerly a CIO for several big-name telcos before joining Check Point in 2019, Yarkoni considers the cloud to be “part of [his] heritage” and one of modern IT’s most valuable tools.
Check Point has three important ‘product families’, Quantum, CloudGuard, and Harmony, with each one providing another layer of holistic IT protection:
- Quantum: secures enterprise networks from sophisticated cyber attacks
- CloudGuard: acts as a scalable and unified cloud-native security platform for the protection of any cloud
- Harmony: protects remote users and devices from cyber threats that might compromise organisational data
However, more than just providing security, Yarkoni emphasises the need for software to be proactive and minimise the possibility of threats in the first instance. This is something Check Point assuredly delivers, “the industry recognises that preventing, not just detecting, is crucial. Check Point has one platform that gives customers the end-to-end cover they need; they don't have to go anywhere else. That level of threat prevention capability is core to our DNA and across all three product lines.”
In many ways, Check Point’s solutions’ capabilities have actually converged to meet the exact working requirements of contemporary enterprise IT. As more companies embark on their own digital transformation journeys in the wake of COVID-19, the inevitability of unforeseen threats increases, which also makes forming security-based partnerships essential. Healthcare of Ontario Pension Plan (HOOPP) sought out Check Point for this very reason when it was in the process of selecting Microsoft Azure as its cloud provider. “Let's be clear: Azure is a secure cloud, but when you operate in a cloud you need several layers of security and governance to prevent mistakes from becoming risks,” Yarkoni clarifies.
The partnership is a distinctly three-way split, with each bringing its own core expertise and competencies. More than that, Check Point, HOOPP and Microsoft are all invested in deepening their understanding of each other at an engineering and developmental level. “Both of our organisations (Check Point and Microsoft) are customer-obsessed: we look at the problem from the eyes of the customer and ask, ‘Are we creating value?’” That kind of focus is proving to be invaluable in the digital era, when the challenges and threats of tomorrow remain unpredictable. In this climate, only the best protected will survive and Check Point is standing by, ready to help.
“HOOPP is an amazing organisation,” concludes Yarkoni. “For us to be successful with a customer and be selected as a partner is actually a badge of honor. It says, ‘We passed a very intense and in-depth inspection by very smart people,’ and for me that’s the best thing about working with organisations like HOOPP.”