Abstract

Despite rapidly increasing numbers of people living with dementia, there are substantial difficulties with diagnosis. Around a third of people living with dementia have not received a diagnosis. Even in people who are diagnosed, there is typically a delay of years between developing symptoms and receiving a diagnosis. These problems mean that people are not empowered by receiving information about what is causing their symptoms, and are unable to access treatments. The first treatment that could slow the progression of dementia has just been approved for use in the USA, making early and accurate diagnosis of dementia an urgent priority.

One important reason for delayed and inaccurate dementia diagnosis is that the human eye is not reliably able to tell from brain scans who has early dementia, and who is likely to have a deterioration in their memory in the future. We propose to use artificial intelligence to derive this information from routine brain scans.

AINOSTICS’ technology represents a breakthrough that would provide an automated, and personalised healthcare platform for assisting the clinical diagnosis of dementia using multi-modal imaging and non-imaging data that are already routinely acquired in healthcare and research settings; useful for both the treatment of patients, and importantly, in the development of novel therapeutics.

AINOSTICS’ technology can automatically and intelligently analyse scans to provide sensitive and accurate micro-structural information about key tissue and organ structures then compare this with information from healthy populations to detect the signatures of disease. We intend for AINOSTICS’ software to become a routine part of clinical practice and drug development as the results of our intelligent analysis will provide clinicians, researchers, and imaging centres a convenient and cost-effective means to get reliable, quantitative and objective diagnostic and prognostic data.

For serious global diseases, AINOSTICS’ technology has the potential to save time during patient assessments, accelerate clinical pathways, standardise the quality of care and improve patient outcomes in addition to making important contributions to the development of disease modifying therapeutics.

Aims

One important reason for delayed and inaccurate dementia diagnosis is that the human eye is not reliably able to tell from brain scans who has early dementia, and who is likely to have a deterioration in their memory in the future. We propose to use artificial intelligence to derive this information from routine brain scans.