How personalised healthcare driven by AI is creating better patient outcomes

Author BIOPublished 3 Min Read

Personalised or precision medicine is set to transform healthcare. As an idea it’s been around for a while, but today vastly increased computing power and the possibilities offered by artificial intelligence is driving the development of personalised medicine across the world. So what is it, and why is AI so important?

The NHS describes personalised medicine as ‘a move away from a “one size fits all” approach to the treatment and care of patients with a particular condition, to one which uses new approaches to better manage patients’ health and targets therapies to achieve the best outcomes in the management of a patient’s disease or predisposition to disease.’ It requires having a holistic understanding of individual patients – their genetics, lifestyle and environment – and then being able to predict the best treatments and disease prevention strategies for them.

All this involves analysing vast amounts of data and is only possible because of AI and machine learning. Machines can find patterns and subgroupings in terabytes of data, far beyond the capabilities of us humans. Artificial intelligence in healthcare makes it possible to predict the likelihood of a patient developing a particular disease before it happens or suggest the best possible treatment based on their genes. Utilising artificial intelligence makes a blanket, one-size-fits all approach to healthcare look outdated. Personalised medicine means much more sophisticated strategies in caring for patients and fundamentally better experiences and better outcomes.

Innovating precision medicine means collaboration between key players in healthcare: from across pharma, tech, data, business the academic world. Though the road ahead will be challenging there’s no doubt that the potential opportunities are huge.

Consider the second leading cause of death across the world: cancer. Targeted treatment based on genetic markers has been available for some particular cancers for a number of years. For instance, 40% of colorectal cancers have the KRAS gene mutation. By testing patients for this gene doctors can then rule out particular treatments known to not work in these cases, sparing people from suffering side effects when there will be no benefit to them. Breast cancer is another area where patients are tested for specific markers in order that they are given more effective treatment. The more that can be learned about different cancers, the more these learnings can be applied. On example of how artificial intelligence is powering knowledge can be found at the Jackson Laboratory (JAX) in Maine, where researchers are working with Microsoft on Project Hanover. They are building machine learning tools that can extract relevant information on cancer from research papers – something very difficult for humans given that 4,000 such papers are published every day. The tool will highlight and rank information, so that human curators can then focus on the most relevant findings, adding to a database of genomic data, clinical trials and treatment options. In turn this means oncologists and healthcare providers to find the best drugs for patients with particular genomic mutations.

In London at UCL researchers have applied machine learning to MRI scans from patients with Alzheimer’s disease and dementia. An algorithm known as SuStain has been able to detect different subtypes within groups of patients. It’s hoped that this far greater stratification will result in more effective treatment; as at the moments drug trials fail because they aren’t effective enough statistically against a very wide group of patients, when they may in fact be very effective for a small group, if that group can be identified.

Meanwhile the Cystic Fibrosis Trust has been investing in artificial intelligence in order review data from nearly 10,000 people on the CF register, a total of 2 million data points in a year. The aim is to look for patterns of symptoms and develop risk scores for developing complications. With the help of a £2.5 million grant from the US-based Cystic Fibrosis Foundation the trust is also supporting the further development of SmartCareCF, a digital healthcare monitor for people with CF. By providing daily data the app can now predict the likelihood of a someone’s condition worsening and requiring antibiotics. It will also help those with CF avoid unnecessary hospital and clinic visits; it’s dangerous for people with CF to meet face to face because of the chances of cross-infection.

These are just a few of the ways precision healthcare driven by AI opens up the potential for better disease prevention and treatment for people across the world. Though providing quality healthcare is one of the biggest challenges of our generation, at least we have the possibilities offered by new technologies on our side.

BIOShare article |
BIOShare article |
;