As AI solutions are increasingly adopted in the healthcare sector, Dr Khalid Alyafei – Chief Medical Informatics Officer at Sidra Medicine in Qatar – discusses the benefits and challenges
The growth of Artificial Intelligence (AI) has been exponential in the last decade and all signs point to this advancing in the coming years. Analysis from the International Data Corporation shows that, in the Middle East and Africa alone, spending on cognitive and AI systems is set to grow from 37.5 million in 2017 to over 100 million by 2021 – a growth rate of 32% per year.
Qatar is one of several countries globally that are making AI a major investment priority, and the country is now building a robust AI research ecosystem as part of its national strategy. That is to say making a considered national effort to design and introduce computer-based solutions that can replace or amplify traditionally human tasks.
The application of AI in the healthcare sector is central to this and – thanks to Qatar’s small geographical size and a government-backed centralized healthcare system – its plans have already been fast-tracked.
Sidra Medicine
At Sidra Medicine, a hospital for women and children located in Doha, we’ve seen huge benefits. Our medical team have been using two Da Vinci robots to regularly perform pediatric and gynecological surgeries and it’s clear that robotics can make an enormous difference to how medicine is practiced.
Using the technology, it is possible to perform complex surgical tasks, which require a high degree of precision, via tiny incisions. Compared to traditional open surgery, this minimally invasive approach results in less pain and scarring for patients, as well as shorter recovery times.
Robots of this level of sophistication can additionally be used to provide doctors with magnified 3D images, giving them an unparalleled view of the area undergoing surgery.
Diagnosis using AI
AI has vast potential in diagnosing conditions too, and the international health sector is making headway in detecting diseases like cancer thanks to the advancement of AI. For example, by creating new algorithms to interpret digital pathology imaging, it is possible to detect eosinophils (disease-fighting white blood cells that indicate the body is fighting cancer or parasitic infections). Such discoveries can aid early diagnosis.
New investments around the world will hopefully advance our ability to detect and prevent diseases further. The UK government recently announced an investment of £250 million for the creation of a new NHS AI laboratory, where health experts can work on predicting who may be susceptible to certain diseases. This could bring invaluable new insights in the years to come.
Virtual healthcare assistants
Thanks to AI, the health sector is additionally able to reach people in new ways. Digital solutions such as virtual healthcare assistants and tele-medicine are enabling healthcare providers to deliver off-site consultative services and reach people in remote locations, as well as those who may be unable to easily travel to a medical facility due to their condition.
Such solutions relieve pressures on healthcare professionals. In countries with high internet penetration – such as Qatar, where 95% of the population are online – digitized healthcare has particular promise.
As our understanding of AI develops, its use in the healthcare space will also expand beyond hospitals, healthcare facilities and digital consultations. In Qatar, plans are currently being developed to use AI software to coordinate the vast number of emergency medical responders and resources needed when hosting international events.
AI challenges
However, while the potential of AI is undeniable, introducing it in the healthcare sector also brings challenges.
While some people may be eager to swap in-person healthcare for digital technologies, it is important to remember that not all audiences may adapt to the change as easily. Healthcare apps and virtual assistants are often intuitive for younger people, but a patient of an older generation may find virtual reality challenging and even intimidating.
For AI algorithms to be successful across an entire population, they would also need to account for neurodiversity. People with autism spectrum disorder (ASD), for instance, engage with stimuli in a different way to other people and may require traditional in-person support or a more innovative AI solution to meet their needs.
As we continue to discover new ways to implement AI, we must also remind ourselves of the end users and ask ourselves on a case by case basis whether AI is genuinely the best solution.
AI will undoubtedly help us to realise new possibilities, but our enthusiasm must continue to be tempered. Stringent vetting of AI-based innovations, coupled with robust training for the medical professionals who will be using them, is essential before any solutions should be introduced to the public.
