Artificial intelligence and deep brain stimulation devices are being utilised by teams at Great Ormond Street Hospital (GOSH), with researchers developing a machine learning tool aiming to predict Parkinson’s disease before onset of symptoms, and a clinical trial team implanting a device to block electrical pathways in order to treat epilepsy.
To examine the ability of a machine learning tool to identify patients with Parkinson’s, researchers receiving funding from the NIHR GOSH Biomedical Centre set the tool to analyse a panel of eight blood-based biomarkers where the concentrations were altered in patients with Parkinson’s. Finding that the tool could provide a diagnosis with high accuracy, the researchers then set it to analyse blood from 72 patients with rapid eye movement behaviour disorder (iRPD). Wider research indicates that up to 80 percent of people with iRPD can go on to develop a brain disorder such as Parkinson’s.
According to GOSH, the machine learning tool was able to identify that 79 percent of the iRPD patients “had the same profile as someone with Parkinson’s”. The team followed up with these patients over a 10-year period and found that the “AI predictions have so far matched the clinical conversion rate”. As a result of the test, researchers have been able to predict 16 patients who went on to develop Parkinson’s, seven years before symptom onset.
The follow-up process is continuing for other patients predicted to develop Parkinson’s in order to provide further verification as to the accuracy of the test. From UCL Great Ormond Street Institute of Child Health, senior author of the study Professor Kevin Mills stated that at present, Parkinson’s care looks like “shutting the stable door after the horse has bolted and we need to start experimental treatments before patients develop symptoms”, adding that the team “set out to use state-of-the-art technology to find new and better biomarkers for Parkinson’s disease and develop them into a test that we can translate into any large NHS laboratory.” Dependent on funding, it is hoped that this could be achieved within two years.
Also at GOSH, the results of a clinical trial using deep brain stimulation (DBS) to treat epilepsy have been shared, with the patient’s daytime seizures reportedly decreasing by 80 percent leading to a “significant impact” on quality of life.
The trial saw surgeons mount a device capable of stimulating specific parts of the brain to the patient’s skull and attach it to electrodes deep within the brain, with the aim of reducing seizure activity by blocking electrical pathways. GOSH notes that unlike other DBS devices which are often mounted on the patient’s chest with wires leading up the neck to the brain, this device was mounted on the skull to mitigate against the risk of leads weakening or becoming damaged as the child grows. It is also rechargeable through the use of headphones, does not require surgery to replace it every three to five years, and features the option to optimise towards seizure patterns.
Since getting “switched on” the device has been delivering constant electrical stimulation, with the patient’s family sharing that “quality of life improvement has been invaluable”.
Martin Tisdall, consultant paediatric neurosurgeon at GOSH and honorary associate professor at UCL, said that the team is “excited to build the evidence base to demonstrate the ability of deep brain stimulation to treat paediatric epilepsy and hope in years to come it will be a standard treatment we can offer.”
GOSH in focus
Earlier in the year, we highlighted GOSH’s announcement that it is part of a consortium of six leading European children’s hospitals who will be working together to advance research on children’s health, collaborating on data management to develop new therapies for children and support new technologies such as AI in healthcare for children.
We also explored a report from the trust marking five years since the launch of its Data Research, Innovation and Virtual Environments (DRIVE) unit, detailing how the unit has supported more than 300 data research projects over the last five years.
