Delivering AI to Support Parkinson’s UK During the Coronavirus Pandemic
Parkinson’s UK is Europe’s leading Parkinson’s support and research charity. The organisation raises around £35m p.a. to drive better care, treatments and quality of life for those with Parkinson’s, their friends, family and their carers.
Following a successful project to develop an AI Strategy for Prostate Cancer UK, Andrew Burgess, Founder of Greenhouse Intelligence, was asked to carry out a similar piece of work with Parkinson’s UK.
However, during March 2020, coronavirus impacted the way all organisations, and particularly charities, worked, and so the project pivoted its focus to identifying what AI could do immediately to help Parkinson’s UK cope with the crisis and the fast-changing needs of their community.
The Work
In mid-March of 2020, all charities were hit with a huge funding shortfall as donations dried up overnight. But, at the same time, they had to continue to support their community, which needed the charity’s help more than ever.
Greenhouse pivoted the strategy work to focus on how AI could help the charity immediately. Through discussions with the Parkinson’s team, there were two big challenges that needed to be faced: to understand the immediate needs of their community and to bring in more donations.
Listening to the community
It is crucial that the communications from Parkinson’s UK out to its communities, and the actions taken, reflect as closely as possible the feelings and concerns of those people. During the Covid-19 crisis, these concerns were changing on a very frequent basis as different rules were announced by the Government and the lockdown continued.
Greenhouse recommended that Parkinson’s UK use a solution from a specialist AI vendor to track the topics and themes that the communities were discussing. By using a technique called Comparative Linguistics, they were able to identify how the conversations on the forum, helpline and social media groups were changing on a week-by-week basis.
The team was able to identify a number of different themes, some of which came and went, whilst others persisted. For example, the importance of physiotherapy was a key finding, and the charity’s efforts to provide remote sessions was driven in part by these insights.
Maximising fundraising efforts
The second recommendation was to optimise the fundraising campaigns by using machine learning to identify potential donors better. The charity already used some analytical methods to help them do this, but Greenhouse Intelligence recommended using machine learning to discover new drivers for cohort selection that hadn’t otherwise been considered.
Using the new models, Parkinson’s UK increased their net income on their next campaign by nearly 10%.
Using AI to support Parkinson’s Disease research
Since those initial successes, we have started to investigate how AI can help amplify the research into Parkinson’s Disease, which has been hit both by the contraction of funding but also the challenges of accessing labs to carry out ‘wet’ research.
Greenhouse Intelligence data scientists are using machine learning to identify new insights from the many and varied data sources that the charity has funded. Although it is still early days, the Greenhouse and Parkinson’s UK team have already been able to identify novel insights into the disease progression rates of people with Parkinson’s. We hope to be able to publish our finding in an academic paper later in the year.
There are many other opportunities to apply Artificial Intelligence in researching Parkinson’s Disease, as well as across the charity. The pandemic has brought into sharp focus how this can be achieved quickly and effectively. We’d like to thank everyone at Parkinson’s UK for their help in accomplishing this, especially Jesse Mears, David Dexter James Bunting.
Author:
Andrew Burgess, Strategic Adviser on AI, Founder, Greenhouse Intelligence
You can read all insights from techUK's AI Week here
Katherine Holden
Katherine joined techUK in May 2018 and currently leads the Data Analytics, AI and Digital ID programme.