Thanks to NHS Scotland, researchers at the University of Edinburgh and the University of Dundee will have access to 1.6 million anonymised brain scans to train an AI towards the early detection of dementia.
Willy Gilder, a 71-year-old retired journalist from Scotland, never imagined his life to take such a twist. Four years ago, at 67, Willy’s once vibrant mind started slipping into a cloud of “depression”. He sought medical treatment, but despite the therapies, his mood didn’t lift. Given his age and a 97 out of 100 score on a cognitive test, the hospital’s psychiatrists decided to send him home.
Not long after, Willy moved to Edinburgh where another doctor, equally perplexed by his symptoms, considered an MRI scan of his brain. The images revealed that parts of Willy’s brain were degenerating, possibly hinting at early-stage dementia. Further clinical examinations confirmed the diagnosis of Alzheimer’s disease, which is responsible for 60-80% of all dementia cases. Common symptoms include challenges with memory, apathy, confusion, loss of judgement, and likely depression.
Alzheimer’s is a slow, silent process, marked by the accumulation of faulty proteins that aggregate in the brain, interfering with cellular functions and ultimately leading to irreversible loss of neurons. Such a process may take 10-to-15 years before the first symptoms emerge.
Although there’s no cure for Alzheimer’s disease, early diagnosis allows patients to adopt lifestyle changes and access medications that can slow the disease’s progression. This was evident in the case of Willy, as he echoes “I was diagnosed early. Now I know that staying mentally active, for example, is going to help me”.
In the meantime, near Willy’s house, the groundbreaking project SCANDAN-PIPaRD is underway at the University of Edinburgh, in partnership with the University of Dundee, aiming to change the outlook for dementia patients. Within this project, researchers are developing an in silico medicine-based clinical decision-making tool that uses artificial intelligence (AI) to help radiologists detect early signs of Alzheimer’s and other dementia-related diseases. However, developing such an innovative tool is scientifically challenging on many counts. It requires an immense amount of imaging and clinical data, such as Willy’s. But accessing patient’s data is rather problematic, due to legitimate data safety, privacy, authorisation and interoperability considerations.
The novelty of this project resides in the NHS Scotland’s Public Benefit and Privacy Panel for Health decision to mediate between the public benefit and the privacy implications connected to the use of health data. Thanks to this deliberation, researchers in the SCANDAN project could access close to 1.6 million anonymised brain images collected during routine clinical visits between 2008 and 2018. Such a database literally represents a gold mine for scientists and could tremendously speed up research towards early diagnosis of dementia without disclosing any personal details.
Willy’s diagnosis underscores the critical need for early detection tools, and this whole story exemplifies the essential role and the pathway that medical data plays in advancing biomedical research and healthcare as a whole. This is a pivotal moment in history to unearth, thanks to in silico medicine technologies and real-world data, the puzzle of how diseases can be detected and treated early on for patients like Willy.
References:
