You may not know it, but data science is playing a role in your life in many ways. When you’re on Netflix, data science is giving you the programmes most relevant to your very own viewing habits, and those of millions of others like you. When you search for information using Google, data science methods are being used to process huge amounts of data in order to work out what is relevant to you – providing a unique result each time.
These techniques are making our every-day lives easier. But wouldn’t it be great if we could use similar methods on mental health data in order to return relevant and useful insights? To help us understand different conditions better, to develop better treatments, and ultimately improve care and services?
These questions were explored in a paper released in the Lancet Psychiatry last week, authored by members of the MQ Data Science Group.
And it’s an important focus of our work here at MQ. We are pioneering a new programme of research that will make use of data science methods alongside huge sets of data for mental health. We’re launching a new funding stream in this area, due to be announced next week. And we’re leading the charge in overcoming barriers to research in this field, through the MQ Data Science Group.
Investing in mental health data science
We’re investing in this area because we believe it has the potential to drive forward a step-change in mental health research. Together, these datasets can provide scientists and clinicians with a range of information: patient questionnaire results, medical records, brain structure information, cognitive data, genetic & lifestyle information (nature vs. nurture), and clinical data from huge numbers of people across different studies.
With advancements in technology, it may even be possible to use wearable devices to link physical information (like heart rate or stress hormones) to wellbeing information (like mood, diet and activity) via patient-led apps.
Of course, these aspirations do not come without their challenges – barriers that still need to be overcome through policy change and better integration. And these were highlighted in the Lancet Psychiatry paper too.
Firstly, it can be difficult to recruit participants for data science research. However, studies show that most mental health service users will allow their records to be used in scientific research, provided there is adequate security and anonymity. Patient confidentiality is of greatest importance, and methods are in place to make the data inaccessible to unauthorised individuals and provide patient anonymity.
Furthermore, marrying together the data from different studies and sources means that scientists must actively work together and share information. Here at MQ, we are working hard to push past these boundaries by providing bi-annual Data Science meetings where scientists meet to discuss these issues and their research.
Finally, we must also help the scientists entering these fields of research. These scientists need access to the datasets that will enable them to answer the most important questions relating to mental health. With no new drug treatments over the last 60 years, these data scientists are needed more than ever in order to use existing data to point the way for mental health research.
But most importantly – what’s clear is that these hurdles can be overcome. And that by doing so we can put data-science to work for us, using it’s huge potential to transform mental health.
This blog was developed using resources from the Milto Goulandris Mental Health Intelligence Library
Last updated: 24 February 2017