- Studies increasingly suggest that artificial intelligence is capable of diagnosing certain illnesses long before doctors can.
- Health service providers such as Babylon Health are leading the way in democratising the use of artificial intelligence in healthcare.
- IBM’s supercomputer, Watson, has greatly sped up the diagnosis of a rare form of leukaemia.
- Google-bought technology, DeepMind, has been found to predict acute kidney injury 48 hours ahead of the best current diagnosis methods.
- Rapid developments in the use of robot-assisted surgery and virtual nursing assistants are also expected across the coming years.
Nowhere is speed more paramount than in the field of medical diagnosis. Developing conditions caught at an early stage are significantly more likely to be successfully treated, with chances of recovery diminishing rapidly with time.
Despite this, NHS waiting lists are growing at an exponential rate. For appointments deemed ‘non-urgent’, patients can be left waiting up to 18 weeks for referrals. But what if artificial intelligence (AI) could detect tell-tale signs of disease in less time, more accurately, and at a fraction of the cost? Increasingly, studies are indicating that this could be a not-so-distant reality - and one with potentially life-saving ramifications.
Artificial intelligence in healthcare: 5 real-world examples
Babylon Health is a health service provider offering remote consultations with healthcare professionals on a mobile application. Alongside options to video-chat with human doctors, patients can ‘talk’ about their symptoms to Babylon via chatbot.
Using artificial intelligence programming designed around a doctor’s brain, Babylon then compares the information to known conditions, as well as the patient’s medical history, to provide instant advice on possible diagnoses and common treatments.
Babylon Health claims its artificial intelligence system, which scored 81% in a Membership of the Royal College of General Practitioners (MRCGP) exam, demonstrates a diagnostic ability that is “on-par with human doctors”.
IBM’s Watson and Leukaemia Diagnosis
In January 2015, a patient in her 60s was admitted to a hospital in Japan and diagnosed with acute myeloid leukaemia. Though the chemotherapy she was prescribed successfully attacked the cancer cells, her recovery from post-remission therapy was unusually slow, leading doctors to suspect a different form of leukaemia.
While conventional tests failed to show any sign of it, IBM’s Watson was able to detect over a thousand genetic mutations in her DNA, filtering out those of diagnostic importance from unrelated hereditary characteristics in just 10 minutes.
By cross-referencing the patient’s genetic data with its own database, Watson had accomplished a level of diagnostic detail which human scientists would take weeks to achieve.
DeepMind and Predicting Kidney Injury
Acute kidney injury is a life-threatening condition causing the kidneys to cease normal functioning with immediate effect. Affecting more than 300,000 people in the US each year, it is notoriously difficult to detect.
DeepMind’s team trained its AI on health records from over 700,000 adult patients, a data set too large to be analysed by humans alone. Trawling through information such as medical histories and current vital signs, the AI was able to predict the likelihood of acute kidney injury with 55.8% accuracy, two full days before its occurrence, and 48 hours ahead of existing methods of diagnosis.
In more severe cases, where patients went on to require kidney dialysis, the AI technology was shown to correctly predict acute kidney injury nine times out of ten.
Virtual Nursing Assistants
Think Alexa, but for your hospital bedside. Virtual nursing assistants replicate the typical behaviour of a nurse: assisting patients with their daily routines; helping to answer medical questions; and reminding them to take medications, or go to appointments.
Care Angel’s virtual nurse assistant Angel is a pioneer of the category. The bot enables wellness checks through AI and voice technologies, to ‘drive better medical outcomes at the lowest cost’.
Accenture estimates that virtual nursing assistants could be the second-largest source of annual savings for the U.S. healthcare industry, cutting as much as $20 billion in costs.
Robot-assisted surgeries are also becoming increasingly common in the mainstream medical sphere.
In 2011, Guy’s and St Thomas’ Hospital in London installed a da Vinci Si HD robotic system, which has been in successful operation ever since. From a dual console system at the patient’s side, featuring 4 robotic arms, surgeons can view details extremely closely (up to 10 times the magnification of human vision), and move instruments in many directions and at various angles, improving dexterity and accuracy.
Accenture estimate that AI-enabled, robot-assisted surgery could save the U.S. healthcare industry $40 billion annually by 2026.