AI is improving healthcare for patients, practitioners, and pharmas
This September, a study by the Stanford University School of Medicine and Unanimous AI found that a combination of human and machine intelligence could offer more accurate diagnoses than human doctors alone. Through Swarm AI technology, Unanimous AI maximises human knowledge by applying artificial intelligence to the wisdom of crowds. Here’s the interesting part: the swarm was also more accurate than intelligent software alone by a significant 22 per cent. So, instead of replacing medical professionals, AI can act as a helpful assistant. As well as more accurate diagnoses, how else can AI augment healthcare services?
1) Medical records
Every two years, the amount of available data in the world doubles. Medical professionals need high quality administrative organisation to make sense of mass datasets. Luckily, this is exactly what AI is particularly good at. DeepMind Health, for example, is collaborating with Moorfields Eye Hospital to improve eye treatment by mining data. By 2050, the number of people suffering from sight loss in the UK is expected to double. In its partnership with Moorfields, DeepMind Health has applied machine learning to one million eye scans, searching for conditions that humans might miss.
2) Clinical decision making
IBM has launched Medical Sieve, an algorithm with analytical reasoning capabilities. Medical Sieve helps with decision making in radiology and cardiology, reducing the viewing load of the clinicians that diagnose conditions. Clinicians often suffer from eye fatigue after viewing hundreds of images, which can effect their ability to identify problems. Medical Sieve’s image guided informatics system works alongside them to provide additional clinical knowledge whilst retaining accurate diagnoses.
3) AI powered health assistants
Meet Mabu, Catalia Health‘s personal healthcare assistant. Mabu is powered by the Catalia Health Platform, which is built on a HIPAA (Health Insurance Portability and Accountability Act) compliant system located in the cloud. Mabu uses the platform to provide real time data about the patients it interacts with. This can include medication doses, application, nutrition, and activities. Collecting personal information allows Mabu to engage in tailored conversations, and consequently provide holistic advice to improve care. Assistants like Mabu could be used in hospitals, clinical trials, care homes or for outpatients.
4) Precision medicine
At Disruption Summit Europe 2018, Dr. Koen Kas heralded the creation of human digital twins. These software models could become an artificially intelligent, predictive analogue for health improvement. This will be part of the genomics revolution – a shift that has already begun. Combined with AI, genomics has the potential to completely rewire how we handle health. Canadian company Deep Genomics has developed an AI powered discovery platform that supports the identification and creation of genetic medicines. It searches for potential drug candidates and predicts the alterations they could cause in genetic diseases.
5) Clinical trials procedure
Traditional clinical trials are slow and expensive, taking years to complete and often failing due to participant fall out. This is largely because of legacy processes that make it difficult both for clinics and the people they invite to join the trials. However, there are various ways that AI can improve clinical trials, from matching potential participants with relevant trials to finding meaning in test results. Another avenue being explored by Pfizer is the creation of ingestible AI enabled sensors, to make sure that trial medication is taken at the right time. This accuracy could also be encouraged by the use of AI assistants like Mabu.
AI can make it easier to order and understand medical records, help with clinical decision making, offer direct advice to patients, inform medicine creation and modernise clinical trials. In other words, the technology is a healthcare powerhouse. Despite fears that AI will eventually eliminate the need for human experts, the above examples show that artificial intelligence is an invaluable tool that works alongside professionals. They also demonstrate the variety of medical roles that artificial intelligence can augment… There’s a lot more to it than futuristic robotic surgeons. The more that AI permeates healthcare, the more people will understand about their own health. The challenge, as ever, will be keeping that data secure.
Want to learn more about disruption in healthcare? Sign up for our free newsletter here.