Patients who have memory and thinking problems attend the memory clinic where doctors carry out assessments and send patients for a brain scan which are interpreted by doctors. Sometimes memory and thinking problems might be a sign of dementia, but other times might be caused by other factors such as depression. When looking at the brain scan, it can be hard to spot very small changes in the brain. This makes it difficult for doctors to determine the cause of the memory and thinking problems. In uncertain situations, patients are diagnosed with mild cognitive disorder (MCD). It can take up to three years from symptom onset before doctors can confirm why patients are experiencing memory and thinking problems. To receive an accurate diagnosis, patients would have to be seen by doctors over time. Unfortunately, this follow-up is not always available within the NHS, so many patients are discharged from the memory clinic without a clear answer. AINOSTICS have developed a new artificial intelligence (AI) computer program that can see patterns in brain scans that may be more difficult to notice with the human eye. This computer program has been tested on brain scans obtained for research purposes. AINOSTICS have found that this technology is 92% accurate in predicting whether patients who have memory and thinking problems will develop dementia. However, this computer program remains to be tested in a real-life clinical setting.
“This project will be instrumental in levelling up dementia diagnosis in the NHS. We urgently need new technologies to deliver more timely and accurate diagnosis in memory clinics, and to ensure that access to diagnosis is not influenced by ethnicity, postcode or socio-economic status. We hope this will help prepare the NHS to deliver future treatments for dementia as equitably as possible.”
-Charles Marshall, Chief Investigator
The ABATED study is a partnership between Queen Mary University of London (QMUL) and AINOSTICS. This study is based at the Wolfson Institute of Population Health at QMUL and funded by the National Institute of Health and Care Research (NIHR) which is the governmental medical research organization in the UK. Our study will test how effective this new computer program is by analysing real-world clinical brain scans from patients with MCD who attend the memory clinic. We want to see how useful this technology is in predicting MCD to dementia conversion. In the future, this technology may allow doctors to be more certain of the diagnosis when a patient is first seen at a memory clinic.
Over three years, we aim to recruit 1,200 mild cognitive disorder (MCD) patients from East London memory clinics who have either:
- Recently been diagnosed with MCD (the looking forward group) or
- Have been diagnosed with MCD in the past (the looking back group)
For the looking forward (prospective) group:
- Patients will be invited to complete brief questionnaires, memory and thinking tests, a blood test and asked if we can access their medical records and brain scan.
- After two years we will invite these patients to a similar study visit.
- Using the information collected at the study visit and their medical record we will check to see if our technology can accurately predict MCD to dementia conversion.
For the looking back (retrospective) group:
- Patients who have not opted out of research will be recruited for the 'looking back' group.
- These patients would have been diagnosed with MCD in the past.
- We will analyse patients’ brain scans using our technology and use their medical records to see if our technology is accurate in predicting MCD to dementia conversion.
ABATED Study Groups Diagram [PDF 95KB]
ABATED Study Procedures Diagram [PDF 119KB]
To reduce the time to diagnose dementia for mild cognitive disorder patients
Finding the cause of memory and thinking problems can take a few years. We want to reduce this waiting time so patients can be diagnosed earlier and access the support they need sooner. We hope that, in future, information from our computer program can be used by doctors, together with all the other tests and assessments taken in the memory clinic, to help reach an early and accurate diagnosis. Patients who may not be developing dementia can be reassured earlier on and can be discharged from the memory clinic and seek relevant support elsewhere if needed.
To make dementia diagnosis fair across ethnically diverse populations
Research shows that some ethnic groups and individuals from poorer areas are diagnosed with dementia later and less rigorously. We want to test our AI technology in patients from East London memory clinics who represent an ethnically diverse and under-served population. If our technology is accurate, and if it can be used in the clinical setting in the future, then it can help to provide a fair and accurate diagnosis across diverse groups.
To improve the accuracy of dementia diagnosis
It is difficult to see small changes on a brain scan. Our technology has shown promise in seeing changes in brain scans that might be missed by the human eye. Detecting these changes may lead to a more accurate diagnosis and reduce the wait time for a diagnosis.
To predict dementia development in mild cognitive disorder (MCD) patients
Patients who have been diagnosed with MCD will be recruited for the 'looking forward' group. Patients will be invited to complete brief questionnaires, memory and thinking tests, a blood test and asked if we can access their medical records and brain scan. After two years we will invite these patients to a similar study visit. Using the information collected at the study visit and their medical records, we will check if our technology can accurately predict MCD to dementia conversion.
To determine the accuracy of our AI technology in previously diagnosed MCD patients
The patients who we aim to recruit for the 'looking back' group would have been diagnosed with MCD in the past. At the time of recruitment, these patients may or may not have developed dementia. We want to analyse their brain scans using our technology to see if it can predict MCD to dementia conversion. We will use patients’ medical records to confirm whether our technology was correct.
To involve members of the public in providing feedback on the study design, methods, analysis and sharing our research
Members of the public are an important part of our research. We want to hear your views on how we design, carry out, analyse and share our research with others for example through our annual focus groups. In the meantime, if you have any thoughts, comments or questions please feel free to contact us.
09/05/2022 - ABATED study funding announcement
22/07/2022 - Formation of ABATED Steering Committee
17/10/2022 - Recruitment for the ABATED study initiated at East London Foundation Trust (ELFT)
02/11/2022 - First focus group for ABATED study
14/11/2022 - First ABATED study participant enrolled
08/03/2023 - Recruitment for the ABATED study initiated at North East London Foundation Trust (NELFT)
19/04/2023 - Six-month ABATED study progress review meeting
Where are we on the recruitment timeline?
ABATED Study Recruitment Progress [PDF 65KB]
