David Bates, CEO and PhD, Linus Health, explains how AI can enhance the detection of certain cognitive conditions.
Cognitive disorders such as Alzheimer’s affect more than 55 million people worldwide. Diagnosing these conditions requires an approach similar to cancer, where patients of all ages are screened regularly for the earliest possible detection that can make the difference in intervention and treatment. But the healthcare system isn’t set up this way, and many patients are diagnosed too late in the disorder’s progression.
Unfortunately, brain health is only addressed after symptoms of decline present themselves, but there is a long window for early intervention that current screenings cannot detect. Disorders like Alzheimer’s can begin forming in the brain up to 20 years before symptoms show, but the challenge is detecting it. Without cognitive screenings at every wellness visit, patients may miss out on the early detection that can make the difference in treatment and intervention of disorders like Alzheimer’s and Parkinson’s.
Current cognitive testing methods include PET scans, MRIs, and spinal taps, which are costly and invasive, and often only available at a hospital, rather than at primary care physician offices. Alternatives to these invasive imaging methods are things such as the Montreal Cognitive Assessment (MoCA), the Mini Mental State Exam (MMSE), the MiniCog, or a Clock Drawing Test, which are all completed on pen and paper. Although they are used around the world, they aren’t very sensitive or specific and scoring for the MoCA and MMSE can vary due to administration bias and patient state of mind. Put simply, as the first line of detection for cognitive disorders, pen-and-paper tests lack the sensitivity and precision to be used for early detection. These imprecise, invasive, costly, and time-consuming methods are failing patients, and the industry needs new tools and processes to address the 16 million people with cognitive disorders and take a new approach to brain health that includes screening at all ages to offer the best chance for early detection.
COVID-19’s impact on brain health is bringing a renewed attention on the approach to brain health and this gap in diagnostics. Bringing screening to all ages, rather than just the elderly, is key to the early detection which can provide a more effective and impactful treatment plan for patients. But doing so will require a shift in routine screen schedules at every doctor’s office to make cognitive testing a standard point of care during all wellness checks.
Instead of using antiquated pen-and-paper tests or invasive and expensive hospital testing, modern technology can be used to capture digital biomarkers to aid providers in early diagnosis. Ubiquitous technologies like a tablet or a smartphone are being paired with artificial intelligence to revolutionise brain health and cognitive disorder detection. The mobile devices can capture digital biomarkers such as voice and speech patterns, visuospatial memory, gait and balance, dual-tasking ability, and fine motor control. After hundreds of digital biomarkers are collected digitally, AI and machine learning analysis provides an assessment of the individual’s brain health in just a matter of minutes.
Insights provided via AI analysis of digital biomarkers enable primary care providers to recommend speciality care or make a diagnosis at the earliest onset of decline. This can speed the time to diagnosis, and an approach with AI technology has been proven effective for both early Alzheimer’s and Parksinson’s detection. A newly published study in Neurology by researchers at the Massachusetts General Hospital (MGH) revealed that Linus Health’s DCTclock assessment, a digital version of the classic Clock Drawing Test (CDT), was effective in identifying the beginnings of Alzheimer’s disease pathology in cognitively normal individuals, those with no outward, observable symptoms of dementia.
The same assessment was also found able to distinguish between Parkinson’s Disease cognitive phenotypes. In a study published in the Journal of Parkinson’s Disease, researchers found a higher association of Parkinson’s diagnosis when the clock was drawn slower, and in fewer strokes, compared to a large clock face drawing which had a lower diagnosis rate. Another study published on Parkinsonism & Related Disorders also showed the DCTclock to have greater sensitivity in detecting Parkinson’s compared to healthy control subjects compared to traditional methods.
Traditional cognitive testing has been reserved for the elderly to manage costs and time. By using AI for diagnostics, cognitive health screening can be given to individuals of all ages without burdening care providers. When the tests take less than three minutes compared to an hour for a pen-and-paper test, brain health can be part of the annual wellness visit to give individuals of all ages the best chance against cognitive disorders.
